Red Hat Enterprise Linux 6

Deployment Guide

Deployment, Configuration and Administration of Red Hat Enterprise Linux 6

Edition 5

Jaromír Hradílek

Red Hat Engineering Content Services

Douglas Silas

Red Hat Engineering Content Services

Martin Prpič

Red Hat Engineering Content Services

Stephen Wadeley

Red Hat Engineering Content Services

Eva Kopalová

Red Hat Engineering Content Services

Peter Ondrejka

Red Hat Engineering Content Services

Ella Deon Lackey

Red Hat Engineering Content Services

Tomáš Čapek

Red Hat Engineering Content Services

Petr Kovář

Red Hat Engineering Content Services

Miroslav Svoboda

Red Hat Engineering Content Services

Petr Bokoč

Red Hat Engineering Content Services

Florian Nadge

Red Hat Engineering Content Services

John Ha

Red Hat Engineering Content Services

David O'Brien

Red Hat Engineering Content Services

Michael Hideo

Red Hat Engineering Content Services

Don Domingo

Red Hat Engineering Content Services

Legal Notice

Copyright © 2010–2013 Red Hat, Inc.
This document is licensed by Red Hat under the Creative Commons Attribution-ShareAlike 3.0 Unported License. If you distribute this document, or a modified version of it, you must provide attribution to Red Hat, Inc. and provide a link to the original. If the document is modified, all Red Hat trademarks must be removed.
Red Hat, as the licensor of this document, waives the right to enforce, and agrees not to assert, Section 4d of CC-BY-SA to the fullest extent permitted by applicable law.
Red Hat, Red Hat Enterprise Linux, the Shadowman logo, JBoss, MetaMatrix, Fedora, the Infinity Logo, and RHCE are trademarks of Red Hat, Inc., registered in the United States and other countries.
Linux® is the registered trademark of Linus Torvalds in the United States and other countries.
Java® is a registered trademark of Oracle and/or its affiliates.
XFS® is a trademark of Silicon Graphics International Corp. or its subsidiaries in the United States and/or other countries.
MySQL® is a registered trademark of MySQL AB in the United States, the European Union and other countries.
Node.js® is an official trademark of Joyent. Red Hat Software Collections is not formally related to or endorsed by the official Joyent Node.js open source or commercial project.
The OpenStack® Word Mark and OpenStack Logo are either registered trademarks/service marks or trademarks/service marks of the OpenStack Foundation, in the United States and other countries and are used with the OpenStack Foundation's permission. We are not affiliated with, endorsed or sponsored by the OpenStack Foundation, or the OpenStack community.
All other trademarks are the property of their respective owners.

Abstract

The Deployment Guide documents relevant information regarding the deployment, configuration and administration of Red Hat Enterprise Linux 6. It is oriented towards system administrators with a basic understanding of the system.
Preface
1. Target Audience
2. How to Read this Book
3. Document Conventions
3.1. Typographic Conventions
3.2. Pull-quote Conventions
3.3. Notes and Warnings
4. Feedback
5. Acknowledgments
I. Basic System Configuration
1. Keyboard Configuration
1.1. Changing the Keyboard Layout
1.2. Adding the Keyboard Layout Indicator
1.3. Setting Up a Typing Break
2. Date and Time Configuration
2.1. Date/Time Properties Tool
2.1.1. Date and Time Properties
2.1.2. Network Time Protocol Properties
2.1.3. Time Zone Properties
2.2. Command Line Configuration
2.2.1. Date and Time Setup
2.2.2. Network Time Protocol Setup
3. Managing Users and Groups
3.1. Introduction to Users and Groups
3.1.1. User Private Groups
3.1.2. Shadow Passwords
3.2. Using the User Manager Tool
3.2.1. Viewing Users and Groups
3.2.2. Adding a New User
3.2.3. Adding a New Group
3.2.4. Modifying User Properties
3.2.5. Modifying Group Properties
3.3. Using Command Line Tools
3.3.1. Adding a New User
3.3.2. Adding a New Group
3.3.3. Creating Group Directories
3.4. Additional Resources
3.4.1. Installed Documentation
4. Gaining Privileges
4.1. The su Command
4.2. The sudo Command
4.3. Additional Resources
II. Package Management
5. Registering a System and Managing Subscriptions
5.1. Using Red Hat Subscription Manager Tools
5.1.1. Launching the Red Hat Subscription Manager GUI
5.1.2. Running the subscription-manager Command-Line Tool
5.2. Registering and Unregistering a System
5.2.1. Registering from the GUI
5.2.2. Registering from the Command Line
5.2.3. Unregistering
5.3. Attaching and Removing Subscriptions
5.3.1. Attaching and Removing Subscriptions through the GUI
5.3.2. Attaching and Removing Subscriptions through the Command Line
5.4. Redeeming Vendor Subscriptions
5.4.1. Redeeming Subscriptions through the GUI
5.4.2. Redeeming Subscriptions through the Command Line
5.5. Attaching Subscriptions from a Subscription Asset Manager Activation Key
5.6. Setting Preferences for Systems
5.6.1. Setting Preferences in the UI
5.6.2. Setting Service Levels Through the Command Line
5.6.3. Setting a Preferred Operating System Release Version in the Command Line
5.7. Managing Subscription Expiration and Notifications
6. Yum
6.1. Checking For and Updating Packages
6.1.1. Checking For Updates
6.1.2. Updating Packages
6.1.3. Preserving Configuration File Changes
6.1.4. Upgrading the System Off-line with ISO and Yum
6.2. Packages and Package Groups
6.2.1. Searching Packages
6.2.2. Listing Packages
6.2.3. Displaying Package Information
6.2.4. Installing Packages
6.2.5. Removing Packages
6.3. Working with Transaction History
6.3.1. Listing Transactions
6.3.2. Examining Transactions
6.3.3. Reverting and Repeating Transactions
6.3.4. Completing Transactions
6.3.5. Starting New Transaction History
6.4. Configuring Yum and Yum Repositories
6.4.1. Setting [main] Options
6.4.2. Setting [repository] Options
6.4.3. Using Yum Variables
6.4.4. Viewing the Current Configuration
6.4.5. Adding, Enabling, and Disabling a Yum Repository
6.4.6. Creating a Yum Repository
6.4.7. Working with Yum Cache
6.5. Yum Plug-ins
6.5.1. Enabling, Configuring, and Disabling Yum Plug-ins
6.5.2. Installing Additional Yum Plug-ins
6.5.3. Plug-in Descriptions
6.6. Additional Resources
7. PackageKit
7.1. Updating Packages with Software Update
7.2. Using Add/Remove Software
7.2.1. Refreshing Software Sources (Yum Repositories)
7.2.2. Finding Packages with Filters
7.2.3. Installing and Removing Packages (and Dependencies)
7.2.4. Installing and Removing Package Groups
7.2.5. Viewing the Transaction Log
7.3. PackageKit Architecture
7.4. Additional Resources
III. Networking
8. NetworkManager
8.1. The NetworkManager Daemon
8.2. Interacting with NetworkManager
8.2.1. Connecting to a Network
8.2.2. Configuring New and Editing Existing Connections
8.2.3. Connecting to a Network Automatically
8.2.4. User and System Connections
8.3. Establishing Connections
8.3.1. Establishing a Wired (Ethernet) Connection
8.3.2. Establishing a Wireless Connection
8.3.3. Establishing a Mobile Broadband Connection
8.3.4. Establishing a VPN Connection
8.3.5. Establishing a DSL Connection
8.3.6. Establishing a Bond Connection
8.3.7. Establishing a VLAN Connection
8.3.8. Establishing an IP-over-InfiniBand (IPoIB) Connection
8.3.9. Configuring Connection Settings
8.4. NetworkManager Architecture
9. Network Interfaces
9.1. Network Configuration Files
9.2. Interface Configuration Files
9.2.1. Ethernet Interfaces
9.2.2. Specific ifcfg Options for Linux on System z
9.2.3. Required ifcfg Options for Linux on System z
9.2.4. Channel Bonding Interfaces
9.2.5. Network Bridge
9.2.6. Setting Up 802.1q VLAN Tagging
9.2.7. Alias and Clone Files
9.2.8. Dialup Interfaces
9.2.9. Other Interfaces
9.3. Interface Control Scripts
9.4. Static Routes and the Default Gateway
9.5. Configuring IPv6 Tokenized Interface Identifiers
9.6. Network Function Files
9.7. Ethtool
9.8. Additional Resources
9.8.1. Installed Documentation
9.8.2. Useful Websites
10. Configure SCTP
10.1. Introduction to Streaming Control Transport Protocol (SCTP)
10.1.1. Comparison of TCP and SCTP Handshaking
10.2. Understanding SCTP
10.2.1. Bundled Streams
10.2.2. Partial Reliability
10.2.3. Message Boundary Preservation
10.2.4. Protocol Event Notifications
10.3. When To Use SCTP
10.4. When Not To Use SCTP
10.5. Compare SCTP to Bonding and Network Teaming
10.6. Using SCTP
10.6.1. Check if SCTP is installed
10.6.2. Install SCTP
10.6.3. Configuring SCTP
10.6.4. Tune SCTP
10.6.5. Troubleshooting SCTP
10.7. Additional Resources
10.7.1. Installed Documentation
10.7.2. Useful Websites
IV. Infrastructure Services
11. Services and Daemons
11.1. Configuring the Default Runlevel
11.2. Configuring the Services
11.2.1. Using the Service Configuration Utility
11.2.2. Using the ntsysv Utility
11.2.3. Using the chkconfig Utility
11.3. Running Services
11.3.1. Determining the Service Status
11.3.2. Starting a Service
11.3.3. Stopping a Service
11.3.4. Restarting a Service
11.4. Additional Resources
11.4.1. Installed Documentation
11.4.2. Related Books
12. Configuring Authentication
12.1. Configuring System Authentication
12.1.1. Launching the Authentication Configuration Tool UI
12.1.2. Selecting the Identity Store for Authentication
12.1.3. Configuring Alternative Authentication Features
12.1.4. Configuring Authentication from the Command Line
12.1.5. Using Custom Home Directories
12.2. Using and Caching Credentials with SSSD
12.2.1. About SSSD
12.2.2. Setting up the sssd.conf File
12.2.3. Starting and Stopping SSSD
12.2.4. SSSD and System Services
12.2.5. Configuring Services: NSS
12.2.6. Configuring Services: PAM
12.2.7. Configuring Services: autofs
12.2.8. Configuring Services: sudo
12.2.9. Configuring Services: OpenSSH and Cached Keys
12.2.10. SSSD and Identity Providers (Domains)
12.2.11. Creating Domains: LDAP
12.2.12. Creating Domains: Identity Management (IdM)
12.2.13. Creating Domains: Active Directory
12.2.14. Configuring Domains: Active Directory as an LDAP Provider (Alternative)
12.2.15. Domain Options: Setting Username Formats
12.2.16. Domain Options: Enabling Offline Authentication
12.2.17. Domain Options: Setting Password Expirations
12.2.18. Domain Options: Using DNS Service Discovery
12.2.19. Domain Options: Using IP Addresses in Certificate Subject Names (LDAP Only)
12.2.20. Creating Domains: Proxy
12.2.21. Creating Domains: Kerberos Authentication
12.2.22. Creating Domains: Access Control
12.2.23. Creating Domains: Primary Server and Backup Servers
12.2.24. Installing SSSD Utilities
12.2.25. SSSD and UID and GID Numbers
12.2.26. Creating Local System Users
12.2.27. Seeding Users into the SSSD Cache During Kickstart
12.2.28. Managing the SSSD Cache
12.2.29. Downgrading SSSD
12.2.30. Using NSCD with SSSD
12.2.31. Troubleshooting SSSD
13. OpenSSH
13.1. The SSH Protocol
13.1.1. Why Use SSH?
13.1.2. Main Features
13.1.3. Protocol Versions
13.1.4. Event Sequence of an SSH Connection
13.2. Configuring OpenSSH
13.2.1. Configuration Files
13.2.2. Starting an OpenSSH Server
13.2.3. Requiring SSH for Remote Connections
13.2.4. Using a Key-Based Authentication
13.3. OpenSSH Clients
13.3.1. Using the ssh Utility
13.3.2. Using the scp Utility
13.3.3. Using the sftp Utility
13.4. More Than a Secure Shell
13.4.1. X11 Forwarding
13.4.2. Port Forwarding
13.5. Additional Resources
13.5.1. Installed Documentation
13.5.2. Useful Websites
V. Servers
14. DHCP Servers
14.1. Why Use DHCP?
14.2. Configuring a DHCP Server
14.2.1. Configuration File
14.2.2. Lease Database
14.2.3. Starting and Stopping the Server
14.2.4. DHCP Relay Agent
14.3. Configuring a DHCP Client
14.4. Configuring a Multihomed DHCP Server
14.4.1. Host Configuration
14.5. DHCP for IPv6 (DHCPv6)
14.6. Additional Resources
14.6.1. Installed Documentation
15. DNS Servers
15.1. Introduction to DNS
15.1.1. Nameserver Zones
15.1.2. Nameserver Types
15.1.3. BIND as a Nameserver
15.2. BIND
15.2.1. Configuring the named Service
15.2.2. Editing Zone Files
15.2.3. Using the rndc Utility
15.2.4. Using the dig Utility
15.2.5. Advanced Features of BIND
15.2.6. Common Mistakes to Avoid
15.2.7. Additional Resources
16. Web Servers
16.1. The Apache HTTP Server
16.1.1. New Features
16.1.2. Notable Changes
16.1.3. Updating the Configuration
16.1.4. Running the httpd Service
16.1.5. Editing the Configuration Files
16.1.6. Working with Modules
16.1.7. Setting Up Virtual Hosts
16.1.8. Setting Up an SSL Server
16.1.9. Additional Resources
17. Mail Servers
17.1. Email Protocols
17.1.1. Mail Transport Protocols
17.1.2. Mail Access Protocols
17.2. Email Program Classifications
17.2.1. Mail Transport Agent
17.2.2. Mail Delivery Agent
17.2.3. Mail User Agent
17.3. Mail Transport Agents
17.3.1. Postfix
17.3.2. Sendmail
17.3.3. Fetchmail
17.3.4. Mail Transport Agent (MTA) Configuration
17.4. Mail Delivery Agents
17.4.1. Procmail Configuration
17.4.2. Procmail Recipes
17.5. Mail User Agents
17.5.1. Securing Communication
17.6. Additional Resources
17.6.1. Installed Documentation
17.6.2. Useful Websites
17.6.3. Related Books
18. Directory Servers
18.1. OpenLDAP
18.1.1. Introduction to LDAP
18.1.2. Installing the OpenLDAP Suite
18.1.3. Configuring an OpenLDAP Server
18.1.4. Running an OpenLDAP Server
18.1.5. Configuring a System to Authenticate Using OpenLDAP
18.1.6. Additional Resources
19. File and Print Servers
19.1. Samba
19.1.1. Introduction to Samba
19.1.2. Samba Daemons and Related Services
19.1.3. Connecting to a Samba Share
19.1.4. Configuring a Samba Server
19.1.5. Starting and Stopping Samba
19.1.6. Samba Server Types and the smb.conf File
19.1.7. Samba Security Modes
19.1.8. Samba Account Information Databases
19.1.9. Samba Network Browsing
19.1.10. Samba with CUPS Printing Support
19.1.11. Samba Distribution Programs
19.1.12. Additional Resources
19.2. FTP
19.2.1. The File Transfer Protocol
19.2.2. The vsftpd Server
19.2.3. Files Installed with vsftpd
19.2.4. Starting and Stopping vsftpd
19.2.5. vsftpd Configuration Options
19.2.6. Additional Resources
19.3. Printer Configuration
19.3.1. Starting the Printer Configuration Tool
19.3.2. Starting Printer Setup
19.3.3. Adding a Local Printer
19.3.4. Adding an AppSocket/HP JetDirect printer
19.3.5. Adding an IPP Printer
19.3.6. Adding an LPD/LPR Host or Printer
19.3.7. Adding a Samba (SMB) printer
19.3.8. Selecting the Printer Model and Finishing
19.3.9. Printing a Test Page
19.3.10. Modifying Existing Printers
19.3.11. Additional Resources
20. Configuring NTP Using ntpd
20.1. Introduction to NTP
20.2. NTP Strata
20.3. Understanding NTP
20.4. Understanding the Drift File
20.5. UTC, Timezones, and DST
20.6. Authentication Options for NTP
20.7. Managing the Time on Virtual Machines
20.8. Understanding Leap Seconds
20.9. Understanding the ntpd Configuration File
20.10. Understanding the ntpd Sysconfig File
20.11. Checking if the NTP Daemon is Installed
20.12. Installing the NTP Daemon (ntpd)
20.13. Checking the Status of NTP
20.14. Configure the Firewall to Allow Incoming NTP Packets
20.14.1. Configure the Firewall Using the Graphical Tool
20.14.2. Configure the Firewall Using the Command Line
20.15. Configure ntpdate Servers
20.16. Configure NTP
20.16.1. Configure Access Control to an NTP Service
20.16.2. Configure Rate Limiting Access to an NTP Service
20.16.3. Adding a Peer Address
20.16.4. Adding a Server Address
20.16.5. Adding a Broadcast or Multicast Server Address
20.16.6. Adding a Manycast Client Address
20.16.7. Adding a Broadcast Client Address
20.16.8. Adding a Manycast Server Address
20.16.9. Adding a Multicast Client Address
20.16.10. Configuring the Burst Option
20.16.11. Configuring the iburst Option
20.16.12. Configuring Symmetric Authentication Using a Key
20.16.13. Configuring the Poll Interval
20.16.14. Configuring Server Preference
20.16.15. Configuring the Time-to-Live for NTP Packets
20.16.16. Configuring the NTP Version to Use
20.17. Configuring the Hardware Clock Update
20.18. Configuring Clock Sources
20.19. Additional Resources
20.19.1. Installed Documentation
20.19.2. Useful Websites
21. Configuring PTP Using ptp4l
21.1. Introduction to PTP
21.1.1. Understanding PTP
21.1.2. Advantages of PTP
21.2. Using PTP
21.2.1. Checking for Driver and Hardware Support
21.2.2. Installing PTP
21.2.3. Starting ptp4l
21.3. Specifying a Configuration File
21.4. Using the PTP Management Client
21.5. Synchronizing the Clocks
21.6. Verifying Time Synchronization
21.7. Serving PTP Time With NTP
21.8. Serving NTP Time With PTP
21.9. Improving Accuracy
21.10. Additional Resources
21.10.1. Installed Documentation
21.10.2. Useful Websites
VI. Monitoring and Automation
22. System Monitoring Tools
22.1. Viewing System Processes
22.1.1. Using the ps Command
22.1.2. Using the top Command
22.1.3. Using the System Monitor Tool
22.2. Viewing Memory Usage
22.2.1. Using the free Command
22.2.2. Using the System Monitor Tool
22.3. Viewing CPU Usage
22.3.1. Using the System Monitor Tool
22.4. Viewing Block Devices and File Systems
22.4.1. Using the lsblk Command
22.4.2. Using the blkid Command
22.4.3. Using the findmnt Command
22.4.4. Using the df Command
22.4.5. Using the du Command
22.4.6. Using the System Monitor Tool
22.5. Viewing Hardware Information
22.5.1. Using the lspci Command
22.5.2. Using the lsusb Command
22.5.3. Using the lspcmcia Command
22.5.4. Using the lscpu Command
22.6. Monitoring Performance with Net-SNMP
22.6.1. Installing Net-SNMP
22.6.2. Running the Net-SNMP Daemon
22.6.3. Configuring Net-SNMP
22.6.4. Retrieving Performance Data over SNMP
22.6.5. Extending Net-SNMP
22.7. Additional Resources
22.7.1. Installed Documentation
23. Viewing and Managing Log Files
23.1. Configuring rsyslog
23.1.1. Global Directives
23.1.2. Modules
23.1.3. Rules
23.1.4. rsyslog Command Line Configuration
23.2. Locating Log Files
23.2.1. Configuring logrotate
23.3. Viewing Log Files
23.4. Adding a Log File
23.5. Monitoring Log Files
23.6. Additional Resources
23.6.1. Installed Documentation
23.6.2. Useful Websites
24. Automating System Tasks
24.1. Cron and Anacron
24.1.1. Installing Cron and Anacron
24.1.2. Running the Crond Service
24.1.3. Configuring Anacron Jobs
24.1.4. Configuring Cron Jobs
24.1.5. Controlling Access to Cron
24.1.6. Black and White Listing of Cron Jobs
24.2. At and Batch
24.2.1. Installing At and Batch
24.2.2. Running the At Service
24.2.3. Configuring an At Job
24.2.4. Configuring a Batch Job
24.2.5. Viewing Pending Jobs
24.2.6. Additional Command Line Options
24.2.7. Controlling Access to At and Batch
24.3. Additional Resources
25. Automatic Bug Reporting Tool (ABRT)
25.1. Installing ABRT and Starting its Services
25.2. Using the Graphical User Interface
25.3. Using the Command Line Interface
25.3.1. Viewing Problems
25.3.2. Reporting Problems
25.3.3. Deleting Problems
25.4. Configuring ABRT
25.4.1. ABRT Events
25.4.2. Standard ABRT Installation Supported Events
25.4.3. Event Configuration in ABRT GUI
25.4.4. ABRT Specific Configuration
25.4.5. Configuring ABRT to Detect a Kernel Panic
25.4.6. Automatic Downloads and Installation of Debuginfo Packages
25.4.7. Configuring Automatic Reporting
25.4.8. Uploading and Reporting Using a Proxy Server
25.5. Configuring Centralized Crash Collection
25.5.1. Configuration Steps Required on a Dedicated System
25.5.2. Configuration Steps Required on a Client System
25.5.3. Saving Package Information
25.5.4. Testing ABRT's Crash Detection
26. OProfile
26.1. Overview of Tools
26.2. Configuring OProfile
26.2.1. Specifying the Kernel
26.2.2. Setting Events to Monitor
26.2.3. Separating Kernel and User-space Profiles
26.3. Starting and Stopping OProfile
26.4. Saving Data
26.5. Analyzing the Data
26.5.1. Using opreport
26.5.2. Using opreport on a Single Executable
26.5.3. Getting more detailed output on the modules
26.5.4. Using opannotate
26.6. Understanding /dev/oprofile/
26.7. Example Usage
26.8. OProfile Support for Java
26.8.1. Profiling Java Code
26.9. Graphical Interface
26.10. OProfile and SystemTap
26.11. Additional Resources
26.11.1. Installed Docs
26.11.2. Useful Websites
VII. Kernel, Module and Driver Configuration
27. Manually Upgrading the Kernel
27.1. Overview of Kernel Packages
27.2. Preparing to Upgrade
27.3. Downloading the Upgraded Kernel
27.4. Performing the Upgrade
27.5. Verifying the Initial RAM Disk Image
27.6. Verifying the Boot Loader
27.6.1. Configuring the GRUB Boot Loader
27.6.2. Configuring the OS/400 Boot Loader
27.6.3. Configuring the YABOOT Boot Loader
28. Working with Kernel Modules
28.1. Listing Currently-Loaded Modules
28.2. Displaying Information About a Module
28.3. Loading a Module
28.4. Unloading a Module
28.5. Setting Module Parameters
28.6. Persistent Module Loading
28.7. Specific Kernel Module Capabilities
28.7.1. Using Multiple Ethernet Cards
28.7.2. Using Channel Bonding
28.8. Additional Resources
29. The kdump Crash Recovery Service
29.1. Installing the kdump Service
29.2. Configuring the kdump Service
29.2.1. Configuring the kdump at First Boot
29.2.2. Using the Kernel Dump Configuration Utility
29.2.3. Configuring kdump on the Command Line
29.2.4. Testing the Configuration
29.3. Analyzing the Core Dump
29.3.1. Running the crash Utility
29.3.2. Displaying the Message Buffer
29.3.3. Displaying a Backtrace
29.3.4. Displaying a Process Status
29.3.5. Displaying Virtual Memory Information
29.3.6. Displaying Open Files
29.3.7. Exiting the Utility
29.4. Additional Resources
29.4.1. Installed Documentation
29.4.2. Useful Websites
A. Consistent Network Device Naming
A.1. Affected Systems
A.2. System Requirements
A.3. Enabling and Disabling the Feature
A.4. Notes for Administrators
B. RPM
B.1. RPM Design Goals
B.2. Using RPM
B.2.1. Finding RPM Packages
B.2.2. Installing and Upgrading
B.2.3. Configuration File Changes
B.2.4. Uninstalling
B.2.5. Freshening
B.2.6. Querying
B.2.7. Verifying
B.3. Checking a Package's Signature
B.3.1. Importing Keys
B.3.2. Verifying Signature of Packages
B.4. Practical and Common Examples of RPM Usage
B.5. Additional Resources
B.5.1. Installed Documentation
B.5.2. Useful Websites
B.5.3. Related Books
C. The X Window System
C.1. The X Server
C.2. Desktop Environments and Window Managers
C.2.1. Desktop Environments
C.2.2. Window Managers
C.3. X Server Configuration Files
C.3.1. The Structure of the Configuration
C.3.2. The xorg.conf.d Directory
C.3.3. The xorg.conf File
C.4. Fonts
C.4.1. Adding Fonts to Fontconfig
C.5. Runlevels and X
C.5.1. Runlevel 3
C.5.2. Runlevel 5
C.6. Additional Resources
C.6.1. Installed Documentation
C.6.2. Useful Websites
D. The sysconfig Directory
D.1. Files in the /etc/sysconfig/ Directory
D.1.1. /etc/sysconfig/arpwatch
D.1.2. /etc/sysconfig/authconfig
D.1.3. /etc/sysconfig/autofs
D.1.4. /etc/sysconfig/clock
D.1.5. /etc/sysconfig/dhcpd
D.1.6. /etc/sysconfig/firstboot
D.1.7. /etc/sysconfig/i18n
D.1.8. /etc/sysconfig/init
D.1.9. /etc/sysconfig/ip6tables-config
D.1.10. /etc/sysconfig/keyboard
D.1.11. /etc/sysconfig/ldap
D.1.12. /etc/sysconfig/named
D.1.13. /etc/sysconfig/network
D.1.14. /etc/sysconfig/ntpd
D.1.15. /etc/sysconfig/quagga
D.1.16. /etc/sysconfig/radvd
D.1.17. /etc/sysconfig/samba
D.1.18. /etc/sysconfig/saslauthd
D.1.19. /etc/sysconfig/selinux
D.1.20. /etc/sysconfig/sendmail
D.1.21. /etc/sysconfig/spamassassin
D.1.22. /etc/sysconfig/squid
D.1.23. /etc/sysconfig/system-config-users
D.1.24. /etc/sysconfig/vncservers
D.1.25. /etc/sysconfig/xinetd
D.2. Directories in the /etc/sysconfig/ Directory
D.3. Additional Resources
D.3.1. Installed Documentation
E. The proc File System
E.1. A Virtual File System
E.1.1. Viewing Virtual Files
E.1.2. Changing Virtual Files
E.2. Top-level Files within the proc File System
E.2.1. /proc/buddyinfo
E.2.2. /proc/cmdline
E.2.3. /proc/cpuinfo
E.2.4. /proc/crypto
E.2.5. /proc/devices
E.2.6. /proc/dma
E.2.7. /proc/execdomains
E.2.8. /proc/fb
E.2.9. /proc/filesystems
E.2.10. /proc/interrupts
E.2.11. /proc/iomem
E.2.12. /proc/ioports
E.2.13. /proc/kcore
E.2.14. /proc/kmsg
E.2.15. /proc/loadavg
E.2.16. /proc/locks
E.2.17. /proc/mdstat
E.2.18. /proc/meminfo
E.2.19. /proc/misc
E.2.20. /proc/modules
E.2.21. /proc/mounts
E.2.22. /proc/mtrr
E.2.23. /proc/partitions
E.2.24. /proc/slabinfo
E.2.25. /proc/stat
E.2.26. /proc/swaps
E.2.27. /proc/sysrq-trigger
E.2.28. /proc/uptime
E.2.29. /proc/version
E.3. Directories within /proc/
E.3.1. Process Directories
E.3.2. /proc/bus/
E.3.3. /proc/bus/pci
E.3.4. /proc/driver/
E.3.5. /proc/fs
E.3.6. /proc/irq/
E.3.7. /proc/net/
E.3.8. /proc/scsi/
E.3.9. /proc/sys/
E.3.10. /proc/sysvipc/
E.3.11. /proc/tty/
E.3.12. /proc/PID/
E.4. Using the sysctl Command
E.5. Additional Resources
E.5.1. Installed Documentation
E.5.2. Useful Websites
F. Revision History
Index

Preface

The Deployment Guide contains information on how to customize the Red Hat Enterprise Linux 6 system to fit your needs. If you are looking for a comprehensive, task-oriented guide for configuring and customizing your system, this is the manual for you.
This manual discusses many intermediate topics such as the following:
  • Installing and managing packages using the graphical PackageKit and command line Yum package managers
  • Setting up a network—from establishing an Ethernet connection using NetworkManager to configuring channel bonding interfaces to increase server bandwidth
  • Configuring DHCP, BIND, Apache HTTP Server, Postfix, Sendmail and other enterprise-class servers and software
  • Gathering information about your system, including obtaining user-space crash data with the Automatic Bug Reporting Tool, and kernel-space crash data with kdump
  • Easily working with kernel modules and upgrading the kernel

1. Target Audience

The Deployment Guide assumes you have a basic understanding of the Red Hat Enterprise Linux operating system. If you need help with the installation of this system, refer to the Red Hat Enterprise Linux 6 Installation Guide.

2. How to Read this Book

This manual is divided into the following main categories:
Part I, “Basic System Configuration”
This part covers basic system administration tasks such as keyboard configuration, date and time configuration, managing users and groups, and gaining privileges.
Chapter 1, Keyboard Configuration covers basic keyboard setup. Read this chapter if you need to change the keyboard layout, add the Keyboard Indicator applet to the panel, or enforce a periodic typing brake.
Chapter 2, Date and Time Configuration covers the configuration of the system date and time. Read this chapter if you need to change the date and time.
Chapter 3, Managing Users and Groups covers the management of users and groups in a graphical user interface and on the command line. Read this chapter if you need to manage users and groups on your system, or enable password aging.
Chapter 4, Gaining Privileges documents how to gain administrative privileges. Read this chapter to learn how to use the su and sudo commands.
Part II, “Package Management”
This part focuses on product subscriptions and entitlements, and describes how to manage software packages on Red Hat Enterprise Linux using both Yum and the PackageKit suite of graphical package management tools.
Chapter 5, Registering a System and Managing Subscriptions provides an overview of subscription management in Red Hat Enterprise Linux and the Red Hat Subscription Manager tools which are available. Read this chapter to learn how to register or unregister a system, activate a machine, and handle product subscriptions and entitlements.
Chapter 6, Yum describes the Yum package manager. Read this chapter for information on how to search, install, update, and uninstall packages on the command line.
Chapter 7, PackageKit describes the PackageKit suite of graphical package management tools. Read this chapter for information on how to search, install, update, and uninstall packages using a graphical user interface.
Part III, “Networking”
This part describes how to configure the network on Red Hat Enterprise Linux.
Chapter 8, NetworkManager focuses on NetworkManager, a dynamic network control and configuration system that attempts to keep network devices and connections up and active when they are available. Read this chapter for information on how to run the NetworkManager daemon, and how to interact with it using the corresponding applet for the notification area.
Chapter 9, Network Interfaces explores various interface configuration files, interface control scripts, and network function files located in the /etc/sysconfig/network-scripts/ directory. Read this chapter for information on how to use these files to configure network interfaces.
Chapter 10, Configure SCTP describes the Streaming Control Transport Protocol (SCTP), a message oriented, reliable transport protocol with direct support for multihoming. Read this chapter for information on when and how to make use of SCTP.
Part IV, “Infrastructure Services”
This part provides information on how to configure services and daemons, configure authentication, and enable remote logins.
Chapter 11, Services and Daemons explains the concept of runlevels, and describes how to set the default one. It also covers the configuration of the services to be run in each of these runlevels, and provides information on how to start, stop, and restart a service. Read this chapter to learn how to manage services on your system.
Chapter 12, Configuring Authentication describes how to configure user information retrieval from Lightweight Directory Access Protocol (LDAP), Network Information Service (NIS), and Winbind user account databases, and provides an introduction to the System Security Services Daemon (SSSD). Read this chapter if you need to configure authentication on your system.
Chapter 13, OpenSSH describes how to enable a remote login via the SSH protocol. It covers the configuration of the sshd service, as well as a basic usage of the ssh, scp, sftp client utilities. Read this chapter if you need a remote access to a machine.
Part V, “Servers”
This part discusses various topics related to servers such as how to set up a web server or share files and directories over the network.
Chapter 14, DHCP Servers guides you through the installation of a Dynamic Host Configuration Protocol (DHCP) server and client. Read this chapter if you need to configure DHCP on your system.
Chapter 15, DNS Servers introduces you to Domain Name System (DNS), explains how to install, configure, run, and administer the BIND DNS server. Read this chapter if you need to configure a DNS server on your system.
Chapter 16, Web Servers focuses on the Apache HTTP Server 2.2, a robust, full-featured open source web server developed by the Apache Software Foundation. Read this chapter if you need to configure a web server on your system.
Chapter 17, Mail Servers reviews modern email protocols in use today, and some of the programs designed to send and receive email, including Postfix, Sendmail, Fetchmail, and Procmail. Read this chapter if you need to configure a mail server on your system.
Chapter 18, Directory Servers covers the installation and configuration of OpenLDAP 2.4, an open source implementation of the LDAPv2 and LDAPv3 protocols. Read this chapter if you need to configure a directory server on your system.
Chapter 19, File and Print Servers guides you through the installation and configuration of Samba, an open source implementation of the Server Message Block (SMB) protocol, and vsftpd, the primary FTP server shipped with Red Hat Enterprise Linux. Additionally, it explains how to use the Printer Configuration tool to configure printers. Read this chapter if you need to configure a file or print server on your system.
Chapter 20, Configuring NTP Using ntpd covers the configuration of the Network Time Protocol (NTP) daemon (ntpd) for updating the system clock. Read this chapter if you need to configure the system to synchronize the clock with a remote Network Time Protocol (NTP) server.
Chapter 21, Configuring PTP Using ptp4l covers the configuration of the Precision Time Protocol (PTP) for updating the system clock. Read this chapter if you need to configure the system to synchronize the clock with a Precision Time Protocol (PTP) server.
Part VI, “Monitoring and Automation”
This part describes various tools that allow system administrators to monitor system performance, automate system tasks, and report bugs.
Chapter 22, System Monitoring Tools discusses applications and commands that can be used to retrieve important information about the system. Read this chapter to learn how to gather essential system information.
Chapter 23, Viewing and Managing Log Files describes the configuration of the rsyslog daemon, and explains how to locate, view, and monitor log files. Read this chapter to learn how to work with log files.
Chapter 24, Automating System Tasks provides an overview of the cron, at, and batch utilities. Read this chapter to learn how to use these utilities to perform automated tasks.
Chapter 25, Automatic Bug Reporting Tool (ABRT) concentrates on ABRT, a system service and a set of tools to collect crash data and send a report to the relevant issue tracker. Read this chapter to learn how to use ABRT on your system.
Chapter 26, OProfile covers OProfile, a low overhead, system-wide performance monitoring tool. Read this chapter for information on how to use OProfile on your system.
Part VII, “Kernel, Module and Driver Configuration”
This part covers various tools that assist administrators with kernel customization.
Chapter 27, Manually Upgrading the Kernel provides important information on how to manually update a kernel package using the rpm command instead of yum. Read this chapter if you cannot update a kernel package with the Yum package manager.
Chapter 28, Working with Kernel Modules explains how to display, query, load, and unload kernel modules and their dependencies, and how to set module parameters. Additionally, it covers specific kernel module capabilities such as using multiple Ethernet cards and using channel bonding. Read this chapter if you need to work with kernel modules.
Chapter 29, The kdump Crash Recovery Service explains how to configure, test, and use the kdump service in Red Hat Enterprise Linux, and provides a brief overview of how to analyze the resulting core dump using the crash debugging utility. Read this chapter to learn how to enable kdump on your system.
Appendix A, Consistent Network Device Naming
This appendix covers consistent network device naming for network interfaces, a feature that changes the name of network interfaces on a system in order to make locating and differentiating the interfaces easier. Read this appendix to learn more about this feature and how to enable or disable it.
Appendix B, RPM
This appendix concentrates on the RPM Package Manager (RPM), an open packaging system used by Red Hat Enterprise Linux, and the use of the rpm utility. Read this appendix if you need to use rpm instead of yum.
Appendix C, The X Window System
This appendix covers the configuration of the X Window System, the graphical environment used by Red Hat Enterprise Linux. Read this appendix if you need to adjust the configuration of your X Window System.
Appendix D, The sysconfig Directory
This appendix outlines some of the files and directories located in the /etc/sysconfig/ directory. Read this appendix if you want to learn more about these files and directories, their function, and their contents.
Appendix E, The proc File System
This appendix explains the concept of a virtual file system, and describes some of the top-level files and directories within the proc file system (that is, the /proc/ directory). Read this appendix if you want to learn more about this file system.

3. Document Conventions

This manual uses several conventions to highlight certain words and phrases and draw attention to specific pieces of information.
In PDF and paper editions, this manual uses typefaces drawn from the Liberation Fonts set. The Liberation Fonts set is also used in HTML editions if the set is installed on your system. If not, alternative but equivalent typefaces are displayed. Note: Red Hat Enterprise Linux 5 and later include the Liberation Fonts set by default.

3.1. Typographic Conventions

Four typographic conventions are used to call attention to specific words and phrases. These conventions, and the circumstances they apply to, are as follows.
Mono-spaced Bold
Used to highlight system input, including shell commands, file names and paths. Also used to highlight keys and key combinations. For example:
To see the contents of the file my_next_bestselling_novel in your current working directory, enter the cat my_next_bestselling_novel command at the shell prompt and press Enter to execute the command.
The above includes a file name, a shell command and a key, all presented in mono-spaced bold and all distinguishable thanks to context.
Key combinations can be distinguished from an individual key by the plus sign that connects each part of a key combination. For example:
Press Enter to execute the command.
Press Ctrl+Alt+F2 to switch to a virtual terminal.
The first example highlights a particular key to press. The second example highlights a key combination: a set of three keys pressed simultaneously.
If source code is discussed, class names, methods, functions, variable names and returned values mentioned within a paragraph will be presented as above, in mono-spaced bold. For example:
File-related classes include filesystem for file systems, file for files, and dir for directories. Each class has its own associated set of permissions.
Proportional Bold
This denotes words or phrases encountered on a system, including application names; dialog-box text; labeled buttons; check-box and radio-button labels; menu titles and submenu titles. For example:
Choose SystemPreferencesMouse from the main menu bar to launch Mouse Preferences. In the Buttons tab, select the Left-handed mouse check box and click Close to switch the primary mouse button from the left to the right (making the mouse suitable for use in the left hand).
To insert a special character into a gedit file, choose ApplicationsAccessoriesCharacter Map from the main menu bar. Next, choose SearchFind… from the Character Map menu bar, type the name of the character in the Search field and click Next. The character you sought will be highlighted in the Character Table. Double-click this highlighted character to place it in the Text to copy field and then click the Copy button. Now switch back to your document and choose EditPaste from the gedit menu bar.
The above text includes application names; system-wide menu names and items; application-specific menu names; and buttons and text found within a GUI interface, all presented in proportional bold and all distinguishable by context.
Mono-spaced Bold Italic or Proportional Bold Italic
Whether mono-spaced bold or proportional bold, the addition of italics indicates replaceable or variable text. Italics denotes text you do not input literally or displayed text that changes depending on circumstance. For example:
To connect to a remote machine using ssh, type ssh username@domain.name at a shell prompt. If the remote machine is example.com and your username on that machine is john, type ssh john@example.com.
The mount -o remount file-system command remounts the named file system. For example, to remount the /home file system, the command is mount -o remount /home.
To see the version of a currently installed package, use the rpm -q package command. It will return a result as follows: package-version-release.
Note the words in bold italics above: username, domain.name, file-system, package, version and release. Each word is a placeholder, either for text you enter when issuing a command or for text displayed by the system.
Aside from standard usage for presenting the title of a work, italics denotes the first use of a new and important term. For example:
Publican is a DocBook publishing system.

3.2. Pull-quote Conventions

Terminal output and source code listings are set off visually from the surrounding text.
Output sent to a terminal is set in mono-spaced roman and presented thus:
books        Desktop   documentation  drafts  mss    photos   stuff  svn
books_tests  Desktop1  downloads      images  notes  scripts  svgs
Source-code listings are also set in mono-spaced roman but add syntax highlighting as follows:
static int kvm_vm_ioctl_deassign_device(struct kvm *kvm,
                 struct kvm_assigned_pci_dev *assigned_dev)
{
         int r = 0;
         struct kvm_assigned_dev_kernel *match;

         mutex_lock(&kvm->lock);

         match = kvm_find_assigned_dev(&kvm->arch.assigned_dev_head,
                                       assigned_dev->assigned_dev_id);
         if (!match) {
                 printk(KERN_INFO "%s: device hasn't been assigned before, "
                   "so cannot be deassigned\n", __func__);
                 r = -EINVAL;
                 goto out;
         }

         kvm_deassign_device(kvm, match);

         kvm_free_assigned_device(kvm, match);

out:
         mutex_unlock(&kvm->lock);
         return r;
}

3.3. Notes and Warnings

Finally, we use three visual styles to draw attention to information that might otherwise be overlooked.

Note

Notes are tips, shortcuts or alternative approaches to the task at hand. Ignoring a note should have no negative consequences, but you might miss out on a trick that makes your life easier.

Important

Important boxes detail things that are easily missed: configuration changes that only apply to the current session, or services that need restarting before an update will apply. Ignoring a box labeled “Important” will not cause data loss but may cause irritation and frustration.

Warning

Warnings should not be ignored. Ignoring warnings will most likely cause data loss.

4. Feedback

If you find a typographical error in this manual, or if you have thought of a way to make this manual better, we would love to hear from you! Please submit a report in Bugzilla against the product Red Hat Enterprise Linux 6.
When submitting a bug report, be sure to provide the following information:
  • Manual's identifier: doc-Deployment_Guide
  • Version number: 6
If you have a suggestion for improving the documentation, try to be as specific as possible when describing it. If you have found an error, please include the section number and some of the surrounding text so we can find it easily.

5. Acknowledgments

Certain portions of this text first appeared in the Deployment Guide, copyright © 2007 Red Hat, Inc., available at https://access.redhat.com/site/documentation/en-US/Red_Hat_Enterprise_Linux/5/html/Deployment_Guide/index.html.
Section 22.6, “Monitoring Performance with Net-SNMP” is based on an article written by Michael Solberg.
The authors of this book would like to thank the following people for their valuable contributions: Adam Tkáč, Andrew Fitzsimon, Andrius Benokraitis, Brian Cleary Edward Bailey, Garrett LeSage, Jeffrey Fearn, Joe Orton, Joshua Wulf, Karsten Wade, Lucy Ringland, Marcela Mašláňová, Mark Johnson, Michael Behm, Miroslav Lichvár, Radek Vokál, Rahul Kavalapara, Rahul Sundaram, Sandra Moore, Zbyšek Mráz, Jan Včelák, Peter Hutterer and James Antill, among many others.

Part I. Basic System Configuration

Chapter 1. Keyboard Configuration

This chapter describes how to change the keyboard layout, as well as how to add the Keyboard Indicator applet to the panel. It also covers the option to enforce a typing break, and explains both advantages and disadvantages of doing so.

1.1. Changing the Keyboard Layout

The installation program has allowed you to configure a keyboard layout for your system. However, the default settings may not always suit your current needs. To configure a different keyboard layout after the installation, use the Keyboard Preferences tool.
To open Keyboard Layout Preferences, select SystemPreferencesKeyboard from the panel, and click the Layouts tab.
Keyboard Layout Preferences

Figure 1.1. Keyboard Layout Preferences


You will be presented with a list of available layouts. To add a new one, click the Add... button below the list, and you will be prompted to choose which layout you want to add.
Choosing a layout

Figure 1.2. Choosing a layout


Currently, there are two ways how to choose the keyboard layout: you can either find it by the country it is associated with (the By country tab), or you can select it by language (the By language tab). In either case, first select the desired country or language from the Country or Language pulldown menu, then specify the variant from the Variants menu. The preview of the layout changes immediately. To confirm the selection, click Add.
Selecting the default layout

Figure 1.3. Selecting the default layout


The layout should appear in the list. To make it the default, select the radio button next to its name. The changes take effect immediately. Note that there is a text-entry field at the bottom of the window where you can safely test your settings. Once you are satisfied, click Close to close the window.
Testing the layout

Figure 1.4. Testing the layout


Disable separate layout for each window

By default, changing the keyboard layout affects the active window only. This means that if you change the layout and switch to another window, this window will use the old one, which might be confusing. To turn this behavior off, clear the Separate layout for each window check box.
Doing this has its drawbacks though, as you will no longer be able to choose the default layout by selecting the radio button as shown in Figure 1.3, “Selecting the default layout”. To make the layout the default, simply drag it to the beginning of the list.

1.2. Adding the Keyboard Layout Indicator

If you want to see what keyboard layout you are currently using, or you would like to switch between different layouts with a single mouse click, add the Keyboard Indicator applet to the panel. To do so, right-click the empty space on the main panel, and select the Add to Panel... option from the pulldown menu.
Adding a new applet

Figure 1.5. Adding a new applet


You will be presented with a list of available applets. Scroll through the list (or start typing keyboard into the search field at the top of the window), select Keyboard Indicator, and click the Add button.
Selecting the Keyboard Indicator

Figure 1.6. Selecting the Keyboard Indicator


The applet appears immediately, displaying the shortened name of the country the current layout is associated with. To display the actual variant, hover the pointer over the applet icon.
The Keyboard Indicator applet

Figure 1.7. The Keyboard Indicator applet


1.3. Setting Up a Typing Break

Typing for a long period of time can be not only tiring, but it can also increase the risk of serious health problems, such as carpal tunnel syndrome. One way of preventing this is to configure the system to enforce typing breaks. To do so, select SystemPreferencesKeyboard from the panel, click the Typing Break tab, and select the Lock screen to enforce typing break check box.
Typing Break Properties

Figure 1.8. Typing Break Properties


To increase or decrease the allowed typing time before the break is enforced, click the up or down button next to the Work interval lasts label respectively. You can do the same with the Break interval lasts setting to alter the length of the break itself. Finally, select the Allow postponing of breaks check box if you want to be able to delay the break in case you need to finish the work. The changes take effect immediately.
Taking a break

Figure 1.9. Taking a break


Next time you reach the time limit, you will be presented with a screen advising you to take a break, and a clock displaying the remaining time. If you have enabled it, the Postpone Break button will be located at the bottom right corner of the screen.

Chapter 2. Date and Time Configuration

This chapter covers setting the system date and time in Red Hat Enterprise Linux, both manually and using the Network Time Protocol (NTP), as well as setting the adequate time zone. Two methods are covered: setting the date and time using the Date/Time Properties tool, and doing so on the command line.

2.1. Date/Time Properties Tool

The Date/Time Properties tool allows the user to change the system date and time, to configure the time zone used by the system, and to set up the Network Time Protocol daemon to synchronize the system clock with a time server. Note that to use this application, you must be running the X Window System (see Appendix C, The X Window System for more information on this topic).
To start the tool, select SystemAdministrationDate & Time from the panel, or type the system-config-date command at a shell prompt (e.g., xterm or GNOME Terminal). Unless you are already authenticated, you will be prompted to enter the superuser password.
Authentication Query

Figure 2.1. Authentication Query


2.1.1. Date and Time Properties

As shown in Figure 2.2, “Date and Time Properties”, the Date/Time Properties tool is divided into two separate tabs. The tab containing the configuration of the current date and time is shown by default.
Date and Time Properties

Figure 2.2. Date and Time Properties


To set up your system manually, follow these steps:
  1. Change the current date. Use the arrows to the left and right of the month and year to change the month and year respectively. Then click inside the calendar to select the day of the month.
  2. Change the current time. Use the up and down arrow buttons beside the Hour, Minute, and Second, or replace the values directly.
Click the OK button to apply the changes and exit the application.

2.1.2. Network Time Protocol Properties

If you prefer an automatic setup, select the check box labeled Synchronize date and time over the network instead. This will display the list of available NTP servers as shown in Figure 2.3, “Network Time Protocol Properties”.
Network Time Protocol Properties

Figure 2.3. Network Time Protocol Properties


Here you can choose one of the predefined servers, edit a predefined server by clicking the Edit button, or add a new server name by clicking Add. In the Advanced Options, you can also select whether you want to speed up the initial synchronization of the system clock, or if you wish to use a local time source.

Note

Your system does not start synchronizing with the NTP server until you click the OK button at the bottom of the window to confirm your changes.
Click the OK button to apply any changes made to the date and time settings and exit the application.

2.1.3. Time Zone Properties

To configure the system time zone, click the Time Zone tab as shown in Figure 2.4, “Time Zone Properties”.
Time Zone Properties

Figure 2.4. Time Zone Properties


There are two common approaches to the time zone selection:
  1. Using the interactive map. Click “zoom in” and “zoom out” buttons next to the map, or click on the map itself to zoom into the selected region. Then choose the city specific to your time zone. A red X appears and the time zone selection changes in the list below the map.
  2. Use the list below the map. To make the selection easier, cities and countries are grouped within their specific continents. Note that non-geographic time zones have also been added to address needs in the scientific community.
If your system clock is set to use UTC, select the System clock uses UTC option. UTC stands for the Universal Time, Coordinated, also known as Greenwich Mean Time (GMT). Other time zones are determined by adding or subtracting from the UTC time.
Click OK to apply the changes and exit the program.

2.2. Command Line Configuration

In case your system does not have the Date/Time Properties tool installed, or the X Window Server is not running, you will have to change the system date and time on the command line. Note that in order to perform actions described in this section, you have to be logged in as a superuser:
~]$ su -
Password:

2.2.1. Date and Time Setup

The date command allows the superuser to set the system date and time manually:
  1. Change the current date. Type the command in the following form at a shell prompt, replacing the YYYY with a four-digit year, MM with a two-digit month, and DD with a two-digit day of the month:
    ~]# date +%D -s YYYY-MM-DD
    For example, to set the date to 2 June 2010, type:
    ~]# date +%D -s 2010-06-02
  2. Change the current time. Use the following command, where HH stands for an hour, MM is a minute, and SS is a second, all typed in a two-digit form:
    ~]# date +%T -s HH:MM:SS
    If your system clock is set to use UTC (Coordinated Universal Time), add the following option:
    ~]# date +%T -s HH:MM:SS -u
    For instance, to set the system clock to 11:26 PM using the UTC, type:
    ~]# date +%T -s 23:26:00 -u
You can check your current settings by typing date without any additional argument:

Example 2.1. Displaying the current date and time

~]$ date
Wed Jun  2 11:58:48 CEST 2010

2.2.2. Network Time Protocol Setup

As opposed to the manual setup described above, you can also synchronize the system clock with a remote server over the Network Time Protocol (NTP). For the one-time synchronization only, use the ntpdate command:
  1. Firstly, check whether the selected NTP server is accessible:
    ~]# ntpdate -q server_address
    For example:
    ~]# ntpdate -q 0.rhel.pool.ntp.org
  2. When you find a satisfactory server, run the ntpdate command followed by one or more server addresses:
    ~]# ntpdate server_address...
    For instance:
    ~]# ntpdate 0.rhel.pool.ntp.org 1.rhel.pool.ntp.org
    Unless an error message is displayed, the system time should now be set. You can check the current by setting typing date without any additional arguments as shown in Section 2.2.1, “Date and Time Setup”.
  3. In most cases, these steps are sufficient. Only if you really need one or more system services to always use the correct time, enable running the ntpdate at boot time:
    ~]# chkconfig ntpdate on
    For more information about system services and their setup, see Chapter 11, Services and Daemons.

    Note

    If the synchronization with the time server at boot time keeps failing, i.e., you find a relevant error message in the /var/log/boot.log system log, try to add the following line to /etc/sysconfig/network:
    NETWORKWAIT=1
However, the more convenient way is to set the ntpd daemon to synchronize the time at boot time automatically:
  1. Open the NTP configuration file /etc/ntp.conf in a text editor such as vi or nano, or create a new one if it does not already exist:
    ~]# nano /etc/ntp.conf
  2. Now add or edit the list of public NTP servers. If you are using Red Hat Enterprise Linux 6, the file should already contain the following lines, but feel free to change or expand these according to your needs:
    server 0.rhel.pool.ntp.org iburst
    server 1.rhel.pool.ntp.org iburst
    server 2.rhel.pool.ntp.org iburst
    server 3.rhel.pool.ntp.org iburst
    
    The iburst directive at the end of each line is to speed up the initial synchronization. As of Red Hat Enterprise Linux 6.5 it is added by default. If upgrading from a previous minor release, and your /etc/ntp.conf file has been modified, then the upgrade to Red Hat Enterprise Linux 6.5 will create a new file /etc/ntp.conf.rpmnew and will not alter the existing /etc/ntp.conf file.
  3. Once you have the list of servers complete, in the same file, set the proper permissions, giving the unrestricted access to localhost only:
    restrict default kod nomodify notrap nopeer noquery
    restrict -6 default kod nomodify notrap nopeer noquery
    restrict 127.0.0.1
    restrict -6 ::1
  4. Save all changes, exit the editor, and restart the NTP daemon:
    ~]# service ntpd restart
  5. Make sure that ntpd is started at boot time:
    ~]# chkconfig ntpd on

Chapter 3. Managing Users and Groups

The control of users and groups is a core element of Red Hat Enterprise Linux system administration. This chapter explains how to add, manage, and delete users and groups in the graphical user interface and on the command line, and covers advanced topics, such as enabling password aging or creating group directories.

3.1. Introduction to Users and Groups

While users can be either people (meaning accounts tied to physical users) or accounts which exist for specific applications to use, groups are logical expressions of organization, tying users together for a common purpose. Users within a group can read, write, or execute files owned by that group.
Each user is associated with a unique numerical identification number called a user ID (UID). Likewise, each group is associated with a group ID (GID). A user who creates a file is also the owner and group owner of that file. The file is assigned separate read, write, and execute permissions for the owner, the group, and everyone else. The file owner can be changed only by root, and access permissions can be changed by both the root user and file owner.
Additionally, Red Hat Enterprise Linux supports access control lists (ACLs) for files and directories which allow permissions for specific users outside of the owner to be set. For more information about this feature, refer to the Access Control Lists chapter of the Storage Administration Guide.

3.1.1. User Private Groups

Red Hat Enterprise Linux uses a user private group (UPG) scheme, which makes UNIX groups easier to manage. A user private group is created whenever a new user is added to the system. It has the same name as the user for which it was created and that user is the only member of the user private group.
User private groups make it safe to set default permissions for a newly created file or directory, allowing both the user and the group of that user to make modifications to the file or directory.
The setting which determines what permissions are applied to a newly created file or directory is called a umask and is configured in the /etc/bashrc file. Traditionally on UNIX systems, the umask is set to 022, which allows only the user who created the file or directory to make modifications. Under this scheme, all other users, including members of the creator's group, are not allowed to make any modifications. However, under the UPG scheme, this group protection is not necessary since every user has their own private group.

3.1.2. Shadow Passwords

In environments with multiple users, it is very important to use shadow passwords provided by the shadow-utils package to enhance the security of system authentication files. For this reason, the installation program enables shadow passwords by default.
The following is a list of the advantages shadow passwords have over the traditional way of storing passwords on UNIX-based systems:
  • Shadow passwords improve system security by moving encrypted password hashes from the world-readable /etc/passwd file to /etc/shadow, which is readable only by the root user.
  • Shadow passwords store information about password aging.
  • Shadow passwords allow the /etc/login.defs file to enforce security policies.
Most utilities provided by the shadow-utils package work properly whether or not shadow passwords are enabled. However, since password aging information is stored exclusively in the /etc/shadow file, any commands which create or modify password aging information do not work. The following is a list of utilities and commands that do not work without first enabling shadow passwords:
  • The chage utility.
  • The gpasswd utility.
  • The usermod command with the -e or -f option.
  • The useradd command with the -e or -f option.

3.2. Using the User Manager Tool

The User Manager application allows you to view, modify, add, and delete local users and groups in the graphical user interface. To start the application, either select SystemAdministrationUsers and Groups from the panel, or type system-config-users at a shell prompt. Note that unless you have superuser privileges, the application will prompt you to authenticate as root.

3.2.1. Viewing Users and Groups

The main window of the User Manager is divided into two tabs: The Users tab provides a list of local users along with additional information about their user ID, primary group, home directory, login shell, and full name. The Groups tab provides a list of local groups with information about their group ID and group members.
Viewing users and groups

Figure 3.1. Viewing users and groups


To find a specific user or group, type the first few letters of the name in the Search filter field and either press Enter, or click the Apply filter button. You can also sort the items according to any of the available columns by clicking the column header.
Red Hat Enterprise Linux reserves user and group IDs below 500 for system users and groups. By default, the User Manager does not display the system users. To view all users and groups, select EditPreferences to open the Preferences dialog box, and clear the Hide system users and groups checkbox.

3.2.2. Adding a New User

To add a new user, click the Add User button. A window as shown in Figure 3.2, “Adding a new user” appears.
Adding a new user

Figure 3.2. Adding a new user


The Add New User dialog box allows you to provide information about the newly created user. In order to create a user, enter the username and full name in the appropriate fields and then type the user's password in the Password and Confirm Password fields. The password must be at least six characters long.

Password security advice

It is advisable to use a much longer password, as this makes it more difficult for an intruder to guess it and access the account without permission. It is also recommended that the password not be based on a dictionary term: use a combination of letters, numbers and special characters.
The Login Shell pulldown list allows you to select a login shell for the user. If you are not sure which shell to select, accept the default value of /bin/bash.
By default, the User Manager application creates the home directory for a new user in /home/username/. You can choose not to create the home directory by clearing the Create home directory checkbox, or change this directory by editing the content of the Home Directory text box. Note that when the home directory is created, default configuration files are copied into it from the /etc/skel/ directory.
Red Hat Enterprise Linux uses a user private group (UPG) scheme. Whenever you create a new user, a unique group with the same name as the user is created by default. If you do not want to create this group, clear the Create a private group for the user checkbox.
To specify a user ID for the user, select Specify user ID manually. If the option is not selected, the next available user ID above 500 is assigned to the new user. Because Red Hat Enterprise Linux reserves user IDs below 500 for system users, it is not advisable to manually assign user IDs 1–499.
Clicking the OK button creates the new user. To configure more advanced user properties, such as password expiration, modify the user's properties after adding the user.

3.2.3. Adding a New Group

To add a new user group, select Add Group from the toolbar. A window similar to Figure 3.3, “New Group” appears. Type the name of the new group. To specify a group ID for the new group, select Specify group ID manually and select the GID. Note that Red Hat Enterprise Linux also reserves group IDs lower than 500 for system groups.
New Group

Figure 3.3. New Group


Click OK to create the group. The new group appears in the group list.

3.2.4. Modifying User Properties

To view the properties of an existing user, click on the Users tab, select the user from the user list, and click Properties from the menu (or choose FileProperties from the pulldown menu). A window similar to Figure 3.4, “User Properties” appears.
User Properties

Figure 3.4. User Properties


The User Properties window is divided into multiple tabbed pages:
  • User Data — Shows the basic user information configured when you added the user. Use this tab to change the user's full name, password, home directory, or login shell.
  • Account Info — Select Enable account expiration if you want the account to expire on a certain date. Enter the date in the provided fields. Select Local password is locked to lock the user account and prevent the user from logging into the system.
  • Password Info — Displays the date that the user's password last changed. To force the user to change passwords after a certain number of days, select Enable password expiration and enter a desired value in the Days before change required: field. The number of days before the user's password expires, the number of days before the user is warned to change passwords, and days before the account becomes inactive can also be changed.
  • Groups — Allows you to view and configure the Primary Group of the user, as well as other groups that you want the user to be a member of.

3.2.5. Modifying Group Properties

To view the properties of an existing group, select the group from the group list and click Properties from the menu (or choose FileProperties from the pulldown menu). A window similar to Figure 3.5, “Group Properties” appears.
Group Properties

Figure 3.5. Group Properties


The Group Users tab displays which users are members of the group. Use this tab to add or remove users from the group. Click OK to save your changes.

3.3. Using Command Line Tools

The easiest way to manage users and groups on Red Hat Enterprise Linux is to use the User Manager application as described in Section 3.2, “Using the User Manager Tool”. However, if you prefer command line tools or do not have the X Window System installed, you can use command line utilities that are listed in Table 3.1, “Command line utilities for managing users and groups”.

Table 3.1. Command line utilities for managing users and groups

Utilities Description
useradd, usermod, userdel Standard utilities for adding, modifying, and deleting user accounts.
groupadd, groupmod, groupdel Standard utilities for adding, modifying, and deleting groups.
gpasswd Standard utility for administering the /etc/group configuration file.
pwck, grpck Utilities that can be used for verification of the password, group, and associated shadow files.
pwconv, pwunconv Utilities that can be used for the conversion of passwords to shadow passwords, or back from shadow passwords to standard passwords.

3.3.1. Adding a New User

To add a new user to the system, typing the following at a shell prompt as root:
useradd [options] username
…where options are command line options as described in Table 3.2, “useradd command line options”.
By default, the useradd command creates a locked user account. To unlock the account, run the following command as root to assign a password:
passwd username
Optionally, you can set password aging policy. Refer to Red Hat Enterprise Linux 6 Security Guide for information on how to enable password aging.

Table 3.2. useradd command line options

Option Description
-c 'comment' comment can be replaced with any string. This option is generally used to specify the full name of a user.
-d home_directory Home directory to be used instead of default /home/username/.
-e date Date for the account to be disabled in the format YYYY-MM-DD.
-f days Number of days after the password expires until the account is disabled. If 0 is specified, the account is disabled immediately after the password expires. If -1 is specified, the account is not be disabled after the password expires.
-g group_name Group name or group number for the user's default group. The group must exist prior to being specified here.
-G group_list List of additional (other than default) group names or group numbers, separated by commas, of which the user is a member. The groups must exist prior to being specified here.
-m Create the home directory if it does not exist.
-M Do not create the home directory.
-N Do not create a user private group for the user.
-p password The password encrypted with crypt.
-r Create a system account with a UID less than 500 and without a home directory.
-s User's login shell, which defaults to /bin/bash.
-u uid User ID for the user, which must be unique and greater than 499.

Explaining the Process

The following steps illustrate what happens if the command useradd juan is issued on a system that has shadow passwords enabled:
  1. A new line for juan is created in /etc/passwd:
    juan:x:501:501::/home/juan:/bin/bash
    The line has the following characteristics:
    • It begins with the username juan.
    • There is an x for the password field indicating that the system is using shadow passwords.
    • A UID greater than 499 is created. Under Red Hat Enterprise Linux, UIDs below 500 are reserved for system use and should not be assigned to users.
    • A GID greater than 499 is created. Under Red Hat Enterprise Linux, GIDs below 500 are reserved for system use and should not be assigned to users.
    • The optional GECOS information is left blank. The GECOS field can be used to provide additional information about the user, such as their full name or phone number.
    • The home directory for juan is set to /home/juan/.
    • The default shell is set to /bin/bash.
  2. A new line for juan is created in /etc/shadow:
    juan:!!:14798:0:99999:7:::
    The line has the following characteristics:
    • It begins with the username juan.
    • Two exclamation marks (!!) appear in the password field of the /etc/shadow file, which locks the account.

      Note

      If an encrypted password is passed using the -p flag, it is placed in the /etc/shadow file on the new line for the user.
    • The password is set to never expire.
  3. A new line for a group named juan is created in /etc/group:
    juan:x:501:
    A group with the same name as a user is called a user private group. For more information on user private groups, refer to Section 3.1.1, “User Private Groups”.
    The line created in /etc/group has the following characteristics:
    • It begins with the group name juan.
    • An x appears in the password field indicating that the system is using shadow group passwords.
    • The GID matches the one listed for user juan in /etc/passwd.
  4. A new line for a group named juan is created in /etc/gshadow:
    juan:!::
    The line has the following characteristics:
    • It begins with the group name juan.
    • An exclamation mark (!) appears in the password field of the /etc/gshadow file, which locks the group.
    • All other fields are blank.
  5. A directory for user juan is created in the /home/ directory:
    ~]# ls -l /home
    total 4
    drwx------. 4 juan juan 4096 Mar  3 18:23 juan
    This directory is owned by user juan and group juan. It has read, write, and execute privileges only for the user juan. All other permissions are denied.
  6. The files within the /etc/skel/ directory (which contain default user settings) are copied into the new /home/juan/ directory:
    ~]# ls -la /home/juan
    total 28
    drwx------. 4 juan juan 4096 Mar  3 18:23 .
    drwxr-xr-x. 5 root root 4096 Mar  3 18:23 ..
    -rw-r--r--. 1 juan juan   18 Jun 22  2010 .bash_logout
    -rw-r--r--. 1 juan juan  176 Jun 22  2010 .bash_profile
    -rw-r--r--. 1 juan juan  124 Jun 22  2010 .bashrc
    drwxr-xr-x. 2 juan juan 4096 Jul 14  2010 .gnome2
    drwxr-xr-x. 4 juan juan 4096 Nov 23 15:09 .mozilla
At this point, a locked account called juan exists on the system. To activate it, the administrator must next assign a password to the account using the passwd command and, optionally, set password aging guidelines.

3.3.2. Adding a New Group

To add a new group to the system, type the following at a shell prompt as root:
groupadd [options] group_name
…where options are command line options as described in Table 3.3, “groupadd command line options”.

Table 3.3. groupadd command line options

Option Description
-f, --force When used with -g gid and gid already exists, groupadd will choose another unique gid for the group.
-g gid Group ID for the group, which must be unique and greater than 499.
-K, --key key=value Override /etc/login.defs defaults.
-o, --non-unique Allow to create groups with duplicate.
-p, --password password Use this encrypted password for the new group.
-r Create a system group with a GID less than 500.

3.3.3. Creating Group Directories

System administrators usually like to create a group for each major project and assign people to the group when they need to access that project's files. With this traditional scheme, file managing is difficult; when someone creates a file, it is associated with the primary group to which they belong. When a single person works on multiple projects, it becomes difficult to associate the right files with the right group. However, with the UPG scheme, groups are automatically assigned to files created within a directory with the setgid bit set. The setgid bit makes managing group projects that share a common directory very simple because any files a user creates within the directory are owned by the group which owns the directory.
For example, a group of people need to work on files in the /opt/myproject/ directory. Some people are trusted to modify the contents of this directory, but not everyone.
  1. As root, create the /opt/myproject/ directory by typing the following at a shell prompt:
    mkdir /opt/myproject
  2. Add the myproject group to the system:
    groupadd myproject
  3. Associate the contents of the /opt/myproject/ directory with the myproject group:
    chown root:myproject /opt/myproject
  4. Allow users to create files within the directory, and set the setgid bit:
    chmod 2775 /opt/myproject
At this point, all members of the myproject group can create and edit files in the /opt/myproject/ directory without the administrator having to change file permissions every time users write new files. To verify that the permissions have been set correctly, run the following command:
~]# ls -l /opt
total 4
drwxrwsr-x. 3 root myproject 4096 Mar  3 18:31 myproject

3.4. Additional Resources

Refer to the following resources for more information about managing users and groups.

3.4.1. Installed Documentation

For information about various utilities for managing users and groups, refer to the following manual pages:
  • chage(1) — A command to modify password aging policies and account expiration.
  • gpasswd(1) — A command to administer the /etc/group file.
  • groupadd(8) — A command to add groups.
  • grpck(8) — A command to verify the /etc/group file.
  • groupdel(8) — A command to remove groups.
  • groupmod(8) — A command to modify group membership.
  • pwck(8) — A command to verify the /etc/passwd and /etc/shadow files.
  • pwconv(8) — A tool to convert standard passwords to shadow passwords.
  • pwunconv(8) — A tool to convert shadow passwords to standard passwords.
  • useradd(8) — A command to add users.
  • userdel(8) — A command to remove users.
  • usermod(8) — A command to modify users.
For information about related configuration files, see:
  • group(5) — The file containing group information for the system.
  • passwd(5) — The file containing user information for the system.
  • shadow(5) — The file containing passwords and account expiration information for the system.

Chapter 4. Gaining Privileges

System administrators (and in some cases users) will need to perform certain tasks with administrative access. Accessing the system as root is potentially dangerous and can lead to widespread damage to the system and data. This chapter covers ways to gain administrative privileges using setuid programs such as su and sudo. These programs allow specific users to perform tasks which would normally be available only to the root user while maintaining a higher level of control and system security.
Refer to the Red Hat Enterprise Linux 6 Security Guide for more information on administrative controls, potential dangers and ways to prevent data loss resulting from improper use of privileged access.

4.1. The su Command

When a user executes the su command, they are prompted for the root password and, after authentication, are given a root shell prompt.
Once logged in via the su command, the user is the root user and has absolute administrative access to the system[1]. In addition, once a user has become root, it is possible for them to use the su command to change to any other user on the system without being prompted for a password.
Because this program is so powerful, administrators within an organization may wish to limit who has access to the command.
One of the simplest ways to do this is to add users to the special administrative group called wheel. To do this, type the following command as root:
usermod -G wheel <username>
In the previous command, replace <username> with the username you want to add to the wheel group.
You can also use the User Manager to modify group memberships, as follows. Note: you need Administrator privileges to perform this procedure.
  1. Click the System menu on the Panel, point to Administration and then click Users and Groups to display the User Manager. Alternatively, type the command system-config-users at a shell prompt.
  2. Click the Users tab, and select the required user in the list of users.
  3. Click Properties on the toolbar to display the User Properties dialog box (or choose Properties on the File menu).
  4. Click the Groups tab, select the check box for the wheel group, and then click OK.
Refer to Section 3.2, “Using the User Manager Tool” for more information about the User Manager.
After you add the desired users to the wheel group, it is advisable to only allow these specific users to use the su command. To do this, you will need to edit the PAM configuration file for su: /etc/pam.d/su. Open this file in a text editor and remove the comment (#) from the following line:
#auth           required        pam_wheel.so use_uid
This change means that only members of the administrative group wheel can switch to another user using the su command.

Note

The root user is part of the wheel group by default.

4.2. The sudo Command

The sudo command offers another approach to giving users administrative access. When trusted users precede an administrative command with sudo, they are prompted for their own password. Then, when they have been authenticated and assuming that the command is permitted, the administrative command is executed as if they were the root user.
The basic format of the sudo command is as follows:
sudo <command>
In the above example, <command> would be replaced by a command normally reserved for the root user, such as mount.
The sudo command allows for a high degree of flexibility. For instance, only users listed in the /etc/sudoers configuration file are allowed to use the sudo command and the command is executed in the user's shell, not a root shell. This means the root shell can be completely disabled as shown in the Red Hat Enterprise Linux 6 Security Guide.
Each successful authentication using the sudo is logged to the file /var/log/messages and the command issued along with the issuer's username is logged to the file /var/log/secure. Should you require additional logging, use the pam_tty_audit module to enable TTY auditing for specified users by adding the following line to your /etc/pam.d/system-auth file:
session required pam_tty_audit.so disable=<pattern> enable=<pattern>
where pattern represents a comma-separated listing of users with an optional use of globs. For example, the following configuration will enable TTY auditing for the root user and disable it for all other users:
session required pam_tty_audit.so disable=* enable=root
Another advantage of the sudo command is that an administrator can allow different users access to specific commands based on their needs.
Administrators wanting to edit the sudo configuration file, /etc/sudoers, should use the visudo command.
To give someone full administrative privileges, type visudo and add a line similar to the following in the user privilege specification section:
juan ALL=(ALL) ALL
This example states that the user, juan, can use sudo from any host and execute any command.
The example below illustrates the granularity possible when configuring sudo:
%users localhost=/sbin/shutdown -h now
This example states that any user can issue the command /sbin/shutdown -h now as long as it is issued from the console.
The man page for sudoers has a detailed listing of options for this file.

Important

There are several potential risks to keep in mind when using the sudo command. You can avoid them by editing the /etc/sudoers configuration file using visudo as described above. Leaving the /etc/sudoers file in its default state gives every user in the wheel group unlimited root access.
  • By default, sudo stores the sudoer's password for a five minute timeout period. Any subsequent uses of the command during this period will not prompt the user for a password. This could be exploited by an attacker if the user leaves his workstation unattended and unlocked while still being logged in. This behavior can be changed by adding the following line to the /etc/sudoers file:
    Defaults    timestamp_timeout=<value>
    where <value> is the desired timeout length in minutes. Setting the <value> to 0 causes sudo to require a password every time.
  • If a sudoer's account is compromised, an attacker can use sudo to open a new shell with administrative privileges:
    sudo /bin/bash
    Opening a new shell as root in this or similar fashion gives the attacker administrative access for a theoretically unlimited amount of time, bypassing the timeout period specified in the /etc/sudoers file and never requiring the attacker to input a password for sudo again until the newly opened session is closed.

4.3. Additional Resources

While programs allowing users to gain administrative privileges are a potential security risk, security itself is beyond the scope of this particular book. You should therefore refer to sources listed below for more information regarding security and privileged access.

Installed Documentation

  • su(1) - the manual page for su provides information regarding the options available with this command.
  • sudo(8) - the manual page for sudo includes a detailed description of this command as well as a list of options available for customizing sudo's behavior.
  • pam(8) - the manual page describing the use of Pluggable Authentication Modules for Linux.

Online Documentation



[1] This access is still subject to the restrictions imposed by SELinux, if it is enabled.

Part II. Package Management

All software on a Red Hat Enterprise Linux system is divided into RPM packages, which can be installed, upgraded, or removed. This part focuses on product subscriptions and entitlements, and describes how to manage packages on Red Hat Enterprise Linux using both Yum and the PackageKit suite of graphical package management tools.

Table of Contents

5. Registering a System and Managing Subscriptions
5.1. Using Red Hat Subscription Manager Tools
5.1.1. Launching the Red Hat Subscription Manager GUI
5.1.2. Running the subscription-manager Command-Line Tool
5.2. Registering and Unregistering a System
5.2.1. Registering from the GUI
5.2.2. Registering from the Command Line
5.2.3. Unregistering
5.3. Attaching and Removing Subscriptions
5.3.1. Attaching and Removing Subscriptions through the GUI
5.3.2. Attaching and Removing Subscriptions through the Command Line
5.4. Redeeming Vendor Subscriptions
5.4.1. Redeeming Subscriptions through the GUI
5.4.2. Redeeming Subscriptions through the Command Line
5.5. Attaching Subscriptions from a Subscription Asset Manager Activation Key
5.6. Setting Preferences for Systems
5.6.1. Setting Preferences in the UI
5.6.2. Setting Service Levels Through the Command Line
5.6.3. Setting a Preferred Operating System Release Version in the Command Line
5.7. Managing Subscription Expiration and Notifications
6. Yum
6.1. Checking For and Updating Packages
6.1.1. Checking For Updates
6.1.2. Updating Packages
6.1.3. Preserving Configuration File Changes
6.1.4. Upgrading the System Off-line with ISO and Yum
6.2. Packages and Package Groups
6.2.1. Searching Packages
6.2.2. Listing Packages
6.2.3. Displaying Package Information
6.2.4. Installing Packages
6.2.5. Removing Packages
6.3. Working with Transaction History
6.3.1. Listing Transactions
6.3.2. Examining Transactions
6.3.3. Reverting and Repeating Transactions
6.3.4. Completing Transactions
6.3.5. Starting New Transaction History
6.4. Configuring Yum and Yum Repositories
6.4.1. Setting [main] Options
6.4.2. Setting [repository] Options
6.4.3. Using Yum Variables
6.4.4. Viewing the Current Configuration
6.4.5. Adding, Enabling, and Disabling a Yum Repository
6.4.6. Creating a Yum Repository
6.4.7. Working with Yum Cache
6.5. Yum Plug-ins
6.5.1. Enabling, Configuring, and Disabling Yum Plug-ins
6.5.2. Installing Additional Yum Plug-ins
6.5.3. Plug-in Descriptions
6.6. Additional Resources
7. PackageKit
7.1. Updating Packages with Software Update
7.2. Using Add/Remove Software
7.2.1. Refreshing Software Sources (Yum Repositories)
7.2.2. Finding Packages with Filters
7.2.3. Installing and Removing Packages (and Dependencies)
7.2.4. Installing and Removing Package Groups
7.2.5. Viewing the Transaction Log
7.3. PackageKit Architecture
7.4. Additional Resources

Chapter 5. Registering a System and Managing Subscriptions

Effective asset management requires a mechanism to handle the software inventory — both the type of products and the number of systems that the software is installed on. The subscription service provides that mechanism and gives transparency into both global allocations of subscriptions for an entire organization and the specific subscriptions assigned to a single system.
Red Hat Subscription Manager works with yum to unite content delivery with subscription management. The Subscription Manager handles only the subscription-system associations. yum or other package management tools handle the actual content delivery. Chapter 6, Yum describes how to use yum.

5.1. Using Red Hat Subscription Manager Tools

Both registration and subscriptions are managed on the local system through GUI and CLI tools called Red Hat Subscription Manager.

Note

The Red Hat Subscription Manager tools are always run as root because of the nature of the changes to the system. However, Red Hat Subscription Manager connects to the subscription service as a user account for the subscription service.

5.1.1. Launching the Red Hat Subscription Manager GUI

Red Hat Subscription Manager is listed as one of the administrative tools in the System > Administration menu in the top management bar.
Red Hat Subscription Manager Menu Option

Figure 5.1. Red Hat Subscription Manager Menu Option


Alternatively, the Red Hat Subscription Manager GUI can be opened from the command line with a single command:
[root@server1 ~]# subscription-manager-gui

5.1.2. Running the subscription-manager Command-Line Tool

Any of the operations that can be performed through the Red Hat Subscription Manager UI can also be performed by running the subscription-manager tool. This tool has the following format:
[root@server1 ~]# subscription-manager command [options]
Each command has its own set of options that are used with it. The subscription-manager help and manpage have more information.

Table 5.1. Common subscription-manager Commands

Command Description
register Registers or identifies a new system to the subscription service.
unregister Unregisters a machine, which strips its subscriptions and removes the machine from the subscription service.
attach Attaches a specific subscription to the machine.
redeem Auto-attaches a machine to a pre-specified subscription that was purchased from a vendor, based on its hardware and BIOS information.
remove Removes a specific subscription or all subscriptions from the machine.
list Lists all of the subscriptions that are compatible with a machine, either subscriptions that are actually attached to the machine or unused subscriptions that are available to the machine.

5.2. Registering and Unregistering a System

Systems can be registered with a subscription service during the firstboot process or as part of the kickstart setup (both described in the Installation Guide). Systems can also be registered after they have been configured or removed from the subscription service inventory (unregistered) if they will no longer be managed within that subscription service.

5.2.1. Registering from the GUI

  1. Launch Subscription Manager. For example:
    [root@server ~]# subscription-manager-gui
  2. If the system is not already registered, then there will be a Register button at the top of the window in the top right corner of the My Installed Products tab.
  3. To identify which subscription server to use for registration, enter the hostname of the service. The default service is Customer Portal Subscription Management, with the hostname subscription.rhn.redhat.com. To use a different subscription service, such as Subscription Asset Manager, enter the hostname of the local server.
    There are seveal different subscription services which use and recognize certificate-based subscriptions, and a system can be registered with any of them in firstboot:
    • Customer Portal Subscription Management, hosted services from Red Hat (the default)
    • Subscription Asset Manager, an on-premise subscription server which proxies content delivery back to the Customer Portal's services
    • CloudForms System Engine, an on-premise service which handles both subscription services and content delivery
  4. Enter the user credentials for the given subscription service to log in.
    The user credentials to use depend on the subscription service. When registering with the Customer Portal, use the Red Hat Network credentials for the administrator or company account.
    However, for Subscription Asset Manager or CloudForms System engine, the user account to use is created within the on-premise service and probably is not the same as the Customer Portal user account.
  5. Optionally, select the Manually assign subscriptions after registration checkbox.
    By default, the registration process automatically attaches the best-matched subscription to the system. This can be turned off so that the subscriptions can be selected manually, as in Section 5.3, “Attaching and Removing Subscriptions”.
  6. When registration begins, Subscription Manager scans for organizations and environments (sub-domains within the organization) to which to register the system.
    IT environments that use Customer Portal Subscription Management have only a single organization, so no further configuration is necessary. IT infrastructures that use a local subscription service like Subscription Asset Manager might have multiple organizations configured, and those organizations may have multiple environments configured within them.
    If multiple organizations are detected, Subscription Manager prompts to select the one to join.
  7. With the default setting, subscriptions are automatically selected and attached to the system. Review and confirm the subscriptions to attach to the system.
    1. If prompted, select the service level to use for the discovered subscriptions.
    2. Subscription Manager lists the selected subscription. This subscription selection must be confirmed by clicking the Subscribe button for the wizard to complete.

5.2.2. Registering from the Command Line

The simplest way to register a machine is to pass the register command with the user account information required to authenticate to Customer Portal Subscription Management. When the system is successfully authenticated, it echoes back the newly-assigned system inventory ID and the user account name which registered it.
The register options are listed in Table 5.2, “register Options”.

Example 5.1. Registering a System to the Customer Portal

[root@server1 ~]# subscription-manager register --username admin-example --password secret

The system has been registered with id: 7d133d55-876f-4f47-83eb-0ee931cb0a97

Example 5.2. Automatically Subscribing While Registering

The register command has an option, --auto-attach, which allows the system to be registered to the subscription service and immediately attaches the subscription which best matches the system's architecture, in a single step.
[root@server1 ~]# subscription-manager register --username admin-example --password secret --auto-attach
This is the same behavior as when registering with the default settings in the Subscription Manager UI.

Example 5.3. Registering a System with Subscription Asset Manager

With Subscription Asset Managr or CloudForms System Engine, an account can have multiple, independent subdivisions called organizationst is required that you specify which organization (essentially an independent group or unit within the main account) to join the system to. This is done by using the --org option in addition to the username and password. The given user must also have the access permissions to add systems to that organization.
To register with a subscription service other than Customer Portal Subscription Management, several additional options must be used to identify the environment and organizational divisions that the system is being registered to:
  • The username and password for the user account withint the subscription service itself
  • --serverurl to give the hostname of the subscription service
  • --baseurl to give the hostname of the content delivery service (for CloudForms System Engine only)
  • --org to give the name of the organization under which to register the system
  • --environment to give the name of an environment (group) within the organization to which to add the system; this is optional, since a default environment is set for any organization
    A system can only be added to an environment during registration.
[root@server1 ~]# subscription-manager register --username=admin-example --password=secret --org="IT Department" --environment="dev" --serverurl=sam-server.example.com

The system has been registered with id: 7d133d55-876f-4f47-83eb-0ee931cb0a97

Note

If the system is in a multi-org environment and no organization is given, the register command returns a Remote Server error.

Table 5.2. register Options

Options Description Required
--username=name Gives the content server user account name. Required
--password=password Gives the password for the user account. Required
--serverurl=hostname Gives the hostname of the subscription service to use. The default is for Customer Portal Subcription Management, subscription.rhn.redhat.com. If this option is not used, the system is registered with Customer Portal Subscription Management. Required for Subscription Asset Manager or CloudForms System Engine
--baseurl=URL Gives the hostname of the content delivery server to use to receive updates. Both Customer Portal Subscription Management and Subscription Asset Manager use Red Hat's hosted content delivery services, with the URL https://cdn.redhat.com. Since CloudForms System Engine hosts its own content, the URL must be used for systems registered with System Engine. Required for CloudForms System Engine
--org=name Gives the organization to which to join the system. Required, except for hosted environments
--environment=name Registers the system to an environment within an organization. Optional
--name=machine_name Sets the name of the system to register. This defaults to be the same as the hostname. Optional
--auto-attach Automatically attaches the best-matched compatible subscription. This is good for automated setup operations, since the system can be configured in a single step. Optional
--activationkey=key Attaches existing subscriptions as part of the registration process. The subscriptions are pre-assigned by a vendor or by a systems administrator using Subscription Asset Manager. Optional
--servicelevel=None|Standard|Premium Sets the service level to use for subscriptions on that machine. This is only used with the --auto-attach option. Optional
--release=NUMBER Sets the operating system minor release to use for subscriptions for the system. Products and updates are limited to that specific minor release version. This is used only used with the --auto-attach option. Optional
--force Registers the system even if it is already registered. Normally, any register operations will fail if the machine is already registered. Optional

5.2.3. Unregistering

The only thing required to unregister a machine is to run the unregister command. This removes the system's entry from the subscription service, removes any subscriptions, and, locally, deletes its identity and subscription certificates.
From the command line, this requires only the unregister command.

Example 5.4. Unregistering a System

[root@server1 ~]# subscription-manager unregister

To unregister from the Subscription Manager GUI:
  1. Open the Subscription Manager UI.
    [root@server ~]# subscription-manager-gui
  2. Open the System menu, and select the Unregister item.
  3. Confirm that the system should be unregistered.

5.3. Attaching and Removing Subscriptions

Assigning a subscription to a system gives the system the ability to install and update any Red Hat product in that subscription. A subscription is a list of all of the products, in all variations, that were purchased at one time, and it defines both the products and the number of times that subscription can be used. When one of those licenses is associated with a system, that subscription is attached to the system.

5.3.1. Attaching and Removing Subscriptions through the GUI

5.3.1.1. Attaching a Subscription

  1. Launch Subscription Manager. For example:
    [root@server ~]# subscription-manager-gui
  2. Open the All Available Subscriptions tab.
  3. Optionally, set the date range and click the Filters button to set the filters to use to search for available subscriptions.
    Subscriptions can be filtered by their active date and by their name. The checkboxes provide more fine-grained filtering:
    • match my system shows only subscriptions which match the system architecture.
    • match my installed products shows subscriptions which work with currently installed products on the system.
    • have no overlap with existing subscriptions excludes subscriptions with duplicate products. If a subscription is already attached to the system for a specific product or if multiple subscriptions supply the same product, then the subscription service filters those subscriptions and shows only the best fit.
    • contain the text searches for strings, such as the product name, within the subscription or pool.
    After setting the date and filters, click the Update button to apply them.
  4. Select one of the available subscriptions.
  5. Click the Subscribe button.

5.3.1.2. Removing Subscriptions

  1. Launch Subscription Manager. For example:
    [root@server ~]# subscription-manager-gui
  2. Open the My Subscriptions tab.
    All of the active subscriptions to which the system is currently attached are listed. (The products available through the subscription may or may not be installed.)
  3. Select the subscription to remove.
  4. Click the Remove button in the bottom right of the window.

5.3.2. Attaching and Removing Subscriptions through the Command Line

5.3.2.1. Attaching Subscriptions

Attaching subscriptions to a system requires specifying the individual product or subscription to attach, using the --pool option.
[root@server1 ~]# subscription-manager attach --pool=XYZ01234567
The options for the attach command are listed in Table 5.3, “attach Options”.
The ID of the subscription pool for the purchased product must be specified. The pool ID is listed with the product subscription information, which is available from running the list command:
[root@server1 ~]# subscription-manager list --available

+-------------------------------------------+
    Available Subscriptions
+-------------------------------------------+
ProductName:            RHEL for Physical Servers
ProductId:              MKT-rhel-server
PoolId:                 ff8080812bc382e3012bc3845ca000cb
Quantity:               10
Expires:                2011-09-20
Alternatively,the best-fitting subscriptions, as identified by the subscription service, can be attached to the system by using the --auto option (which is analogous to the --auto-attach option with the register command).
[root@server1 ~]# subscription-manager attach --auto

Table 5.3. attach Options

Options Description Required
--pool=pool-id Gives the ID for the subscription to attach to the system. Required, unless --auto is used
--auto Automatically attaches the system to the best-match subscription or subscriptions. Optional
--quantity=number Attaches multiple counts of a subscription to the system. This is used to cover subscriptions that define a count limit, like using two 2-socket server subscriptions to cover a 4-socket machine. Optional
--servicelevel=None|Standard|Premium Sets the service level to use for subscriptions on that machine. This is only used with the --auto option. Optional

5.3.2.2. Removing Subscriptions from the Command Line

A system can be attached to multiple subscriptions and products. Similarly, a single subscription or all subscriptions can be removed from the system.
Running the remove command with the --all option removes every product subscription and subscription pool that is currently attached to the system.
[root@server1 ~]# subscription-manager remove --all
It is also possible to remove a single product subscription. Each product has an identifying X.509 certificate installed with it. The product subscription to remove is identified in the remove command by referencing the ID number of that X.509 certificate.
  1. Get the serial number for the product certificate, if you are removing a single product subscription. The serial number can be obtained from the subscription#.pem file (for example, 392729555585697907.pem) or by using the list command. For example:
    [root@server1 ~]# subscription-manager list --consumed
    
    +-------------------------------------------+
        Consumed Product Subscriptions
    +-------------------------------------------+
    
    
    ProductName:         High availability (cluster suite)
    ContractNumber:      0
    SerialNumber:        11287514358600162
    Active:              True
    Begins:              2010-09-18
    Expires:             2011-11-18
  2. Run the subscription-manager tool with the --serial option to specify the certificate.
    [root@server1 ~]# subscription-manager remove --serial=11287514358600162

5.4. Redeeming Vendor Subscriptions

Systems can be set up with pre-existing subscriptions already available to that system. For some systems which were purchased through third-party vendors, a subscription to Red Hat products is included with the purchase of the machine.
Red Hat Subscription Manager pulls information about the system hardware and the BIOS into the system facts to recognize the hardware vendor. If the vendor and BIOS information matches a certain configuration, then the subscription can be redeemed, which will allow subscriptions to be automatically attached to the system.

5.4.1. Redeeming Subscriptions through the GUI

Note

If the machine does not have any subscriptions to be redeemed, then the Redeem menu item is not there.
  1. Launch Subscription Manager. For example:
    [root@server ~]# subscription-manager-gui
  2. If necessary, register the system, as described in Section 5.2.1, “Registering from the GUI”.
  3. Open the System menu in the top left of the window, and click the Redeem item.
  4. In the dialog window, enter the email address to send the notification to when the redemption is complete. Because the redemption process can take several minutes to contact the vendor and receive information about the pre-configured subscriptions, the notification message is sent through email rather than through the Subscription Manager dialog window.
  5. Click the Redeem button.
It can take up to ten minutes for the confirmation email to arrive.

5.4.2. Redeeming Subscriptions through the Command Line

Note

The machine must be registered first so that the subscription service can properly identify the system and its subscriptions.
The machine subscriptions are redeemed by running the redeem command, with an email address to send the redemption email to when the process is complete.
# subscription-manager redeem --email=jsmith@example.com

5.5. Attaching Subscriptions from a Subscription Asset Manager Activation Key

A local Subscription Asset Manager can pre-configure subscriptions to use for a system, and that pre-configured set of subscriptions is identified by an activation key. That key can then be used to attach those subscriptions on a local system.
The Subscription Asset Manager activation key can be used as part of the registration process for the new system:
# subscription-manager register --username=jsmith --password=secret --org="IT Dept" --activationkey=abcd1234
If there are multiple organizations, it is still necessary to specify the organization for the system. That information is not defined in the activation key.

5.6. Setting Preferences for Systems

Auto-attaching and healing (updating) subscriptions selects what subscriptions to attach to a system based on a variety of criteria, including current installed products, hardware, and architecture. It is possible to set two additional preferences for Subscription Manager to use:
  • Service levels for subscriptions
  • The operating system minor version (X.Y) to use
This is especially useful for healing, which runs daily to ensure that all installed products and current subscriptions remain active.

5.6.1. Setting Preferences in the UI

Both a service level preference and an operating system release version preference are set in the System Preferences dialog box in Subscription Manager.
  1. Open the Subscription Manager.
  2. Open the System menu.
  3. Select the System Preferences menu item.
  4. Select the desired service level agreement preference from the drop-down menu. Only service levels available to the Red Hat account, based on all of its active subscriptions, are listed.
  5. Select the operating system release preference in the Release version drop-down menu. The only versions listed are Red Hat Enterprise Linux versions for which the account has an active subscription.
  6. The preferences are saved and applied to future subscription operations when they are set. To close the dialog, click Close.

5.6.2. Setting Service Levels Through the Command Line

A general service level preference can be set using the service-level --set command.

Example 5.5. Setting a Service Level Preference

First, list the available service levels for the system, using the --list option with the service-level command.
[root@server ~]# subscription-manager service-level --list
+-------------------------------------------+
          Available Service Levels
+-------------------------------------------+
Standard
None
Premium
Self-Support
Then, set the desired level for the system.
[root@server ~]# subscription-manager service-level --set=self-support
Service level set to: self-support
The current setting for the local system is shown with the --show option:
[root#server ~]# subscription-manager service-level --show
Current service level: self-support

A service level preference can be defined when a subscription operation is being run (such as registering a system or attaching subscriptions after registration). This can be used to override a system preference. Both the register and attach commands have the --servicelevel option to set a preference for that action.

Example 5.6. Autoattaching Subscriptions with a Premium Service Level

[root#server ~]# subscription-manager attach --auto --servicelevel Premium
Service level set to: Premium
Installed Product Current Status:
ProductName:            RHEL 6 for Workstations
Status:                 Subscribed

Note

The --servicelevel option requires the --auto-attach option (for register) or --auto option (for attach). It cannot be used when attaching a specified pool or when importing a subscription.

5.6.3. Setting a Preferred Operating System Release Version in the Command Line

Many IT environments have to be certified to meet a certain level of security or other criteria. In that case, major upgrades must be carefully planned and controlled — so administrators cannot simply run yum update and move from version to version.
Setting a release version preference limits the system access to content repositories associated with that operating system version instead of automatically using the newest or latest version repositories.
For example, if the preferred operating system version is 6.3, then 6.3 content repositories will be preferred for all installed products and attached subscriptions for the system, even as other repositories become available.

Example 5.7. Setting an Operating System Release During Registration

A preference for a release version can be set when the system is registered by using --release option with the register. This applies the release preference to any subscriptions selected and auto-attached to the system at registration time.
Setting a preference requires the --auto-attach option, because it is one of the criteria used to select subscriptions to auto-attach.
[root#server ~]# subscription-manager register --auto-attach --release=6.4 --username=admin@example.com...

Note

Unlike setting a service level preference, a release preference can only be used during registration or set as a preference. It cannot be specified with the attach command.

Example 5.8. Setting an Operating System Release Preference

The release command can display the available operating system releases, based on the available, purchased (not only attached) subscriptions for the organization.
[root#server ~]# subscription-manager release --list
+-------------------------------------------+
          Available Releases
+-------------------------------------------+
6.2
6.3
The --set then sets the preference to one of the available release versions:
[root#server ~]# subscription-manager release --set=6.3
Release version set to: 6.3

5.7. Managing Subscription Expiration and Notifications

Subscriptions are active for a certain period of time, called the validity period. When a subscription is purchased, the start and end dates for the contract are set.
On a system, there can be multiple subscriptions attached. Each product requires its own subscription. Additionally, some products may require multiple quantities for it to be fully subscribed. For example, a 16 socket machine may require four 4-socket operating system subscriptions to cover the socket count.
The My Installed Software tab shows the subscription status for the entire system. It also shows a date; that is the first date that a product subscription goes from valid to invalid (meaning it expires).
Valid Until...

Figure 5.2. Valid Until...


The Red Hat Subscription Manager provides a series of log and UI messages that indicate any changes to the valid certificates of any installed products for a system. In the Subscription Manager GUI, the status of the system subscriptions is color-coded, where green means all products are fully subscribed, yellow means that some products may not be subscribed but updates are still in effect, and red means that updates are disabled.
Color-Coded Status Views

Figure 5.3. Color-Coded Status Views


The command-line tools also indicate that status of the machine. The green, yellow, and red codes translate to text status messages of subscribed, partially subscribed, and expired/not subscribed, respectively.
[root@server ~]# subscription-manager list
+-------------------------------------------+
    Installed Product Status
+-------------------------------------------+

ProductName:            Red Hat Enterprise Linux Server
Status: Not Subscribed
Expires:                                         
SerialNumber:                                    
ContractNumber:                                  
AccountNumber:
Whenever there is a warning about subscription changes, a small icon appears in the top menu bar, similar to a fuel gauge.
Subscription Notification Icon

Figure 5.4. Subscription Notification Icon


As any installed product nears the expiration date of the subscription, the Subscription Manager daemon will issue a warning. A similar message is given when the system has products without a valid certificate, meaning either a subscription is not attached that covers that product or the product is installed past the expiration of the subscription. Clicking the Manage My Subscriptions... button in the subscription notification window opens the Red Hat Subscription Manager GUI to view and update subscriptions.
Subscription Warning Message

Figure 5.5. Subscription Warning Message


When the Subscription Manager UI opens, whether it was opened through a notification or just opened normally, there is an icon in the upper left corner that shows whether products lack a valid certificate. The easiest way to attach subscriptions which match invalidated products is to click the Autoattach button.
Autoattach Button

Figure 5.6. Autoattach Button


The Subscribe System dialog shows a targeted list of available subscriptions that apply to the specific products that do not have valid certificates (assuming subscriptions are available).
Subscribe System

Figure 5.7. Subscribe System


Chapter 6. Yum

Yum is the Red Hat package manager that is able to query for information about available packages, fetch packages from repositories, install and uninstall them, and update an entire system to the latest available version. Yum performs automatic dependency resolution on packages you are updating, installing, or removing, and thus is able to automatically determine, fetch, and install all available dependent packages.
Yum can be configured with new, additional repositories, or package sources, and also provides many plug-ins which enhance and extend its capabilities. Yum is able to perform many of the same tasks that RPM can; additionally, many of the command line options are similar. Yum enables easy and simple package management on a single machine or on groups of them.

Secure package management with GPG-signed packages

Yum provides secure package management by enabling GPG (Gnu Privacy Guard; also known as GnuPG) signature verification on GPG-signed packages to be turned on for all package repositories (i.e. package sources), or for individual repositories. When signature verification is enabled, Yum will refuse to install any packages not GPG-signed with the correct key for that repository. This means that you can trust that the RPM packages you download and install on your system are from a trusted source, such as Red Hat, and were not modified during transfer. Refer to Section 6.4, “Configuring Yum and Yum Repositories” for details on enabling signature-checking with Yum, or Section B.3, “Checking a Package's Signature” for information on working with and verifying GPG-signed RPM packages in general.
Yum also enables you to easily set up your own repositories of RPM packages for download and installation on other machines.
Learning Yum is a worthwhile investment because it is often the fastest way to perform system administration tasks, and it provides capabilities beyond those provided by the PackageKit graphical package management tools. Refer to Chapter 7, PackageKit for details on using PackageKit.

Yum and superuser privileges

You must have superuser privileges in order to use yum to install, update or remove packages on your system. All examples in this chapter assume that you have already obtained superuser privileges by using either the su or sudo command.

6.1. Checking For and Updating Packages

6.1.1. Checking For Updates

To see which installed packages on your system have updates available, use the following command:
yum check-update
For example:
~]# yum check-update
Loaded plugins: product-id, refresh-packagekit, subscription-manager
Updating Red Hat repositories.
INFO:rhsm-app.repolib:repos updated: 0
PackageKit.x86_64                  0.5.8-2.el6                rhel
PackageKit-glib.x86_64             0.5.8-2.el6                rhel
PackageKit-yum.x86_64              0.5.8-2.el6                rhel
PackageKit-yum-plugin.x86_64       0.5.8-2.el6                rhel
glibc.x86_64                       2.11.90-20.el6             rhel
glibc-common.x86_64                2.10.90-22                 rhel
kernel.x86_64                      2.6.31-14.el6              rhel
kernel-firmware.noarch             2.6.31-14.el6              rhel
rpm.x86_64                         4.7.1-5.el6                rhel
rpm-libs.x86_64                    4.7.1-5.el6                rhel
rpm-python.x86_64                  4.7.1-5.el6                rhel
udev.x86_64                        147-2.15.el6               rhel
yum.noarch                         3.2.24-4.el6               rhel
The packages in the above output are listed as having updates available. The first package in the list is PackageKit, the graphical package manager. The line in the example output tells us:
  • PackageKit — the name of the package
  • x86_64 — the CPU architecture the package was built for
  • 0.5.8 — the version of the updated package to be installed
  • rhel — the repository in which the updated package is located
The output also shows us that we can update the kernel (the kernel package), Yum and RPM themselves (the yum and rpm packages), as well as their dependencies (such as the kernel-firmware, rpm-libs, and rpm-python packages), all using yum.

6.1.2. Updating Packages

You can choose to update a single package, multiple packages, or all packages at once. If any dependencies of the package (or packages) you update have updates available themselves, then they are updated too.

Updating a Single Package

To update a single package, run the following command as root:
yum update package_name
For example, to update the udev package, type:
~]# yum update udev
Loaded plugins: product-id, refresh-packagekit, subscription-manager
Updating Red Hat repositories.
INFO:rhsm-app.repolib:repos updated: 0
Setting up Update Process
Resolving Dependencies
--> Running transaction check
---> Package udev.x86_64 0:147-2.15.el6 set to be updated
--> Finished Dependency Resolution

Dependencies Resolved

===========================================================================
 Package       Arch            Version                 Repository     Size
===========================================================================
Updating:
 udev          x86_64          147-2.15.el6            rhel          337 k

Transaction Summary
===========================================================================
Install       0 Package(s)
Upgrade       1 Package(s)

Total download size: 337 k
Is this ok [y/N]:
This output contains several items of interest:
  1. Loaded plugins: product-id, refresh-packagekit, subscription-manageryum always informs you which Yum plug-ins are installed and enabled. Refer to Section 6.5, “Yum Plug-ins” for general information on Yum plug-ins, or to Section 6.5.3, “Plug-in Descriptions” for descriptions of specific plug-ins.
  2. udev.x86_64 — you can download and install new udev package.
  3. yum presents the update information and then prompts you as to whether you want it to perform the update; yum runs interactively by default. If you already know which transactions the yum command plans to perform, you can use the -y option to automatically answer yes to any questions that yum asks (in which case it runs non-interactively). However, you should always examine which changes yum plans to make to the system so that you can easily troubleshoot any problems that might arise.
    If a transaction does go awry, you can view Yum's transaction history by using the yum history command as described in Section 6.3, “Working with Transaction History”.

Updating and installing kernels with Yum

yum always installs a new kernel in the same sense that RPM installs a new kernel when you use the command rpm -i kernel. Therefore, you do not need to worry about the distinction between installing and upgrading a kernel package when you use yum: it will do the right thing, regardless of whether you are using the yum update or yum install command.
When using RPM, on the other hand, it is important to use the rpm -i kernel command (which installs a new kernel) instead of rpm -u kernel (which replaces the current kernel). Refer to Section B.2.2, “Installing and Upgrading” for more information on installing/upgrading kernels with RPM.

Updating All Packages and Their Dependencies

To update all packages and their dependencies, simply enter yum update (without any arguments):
yum update

Updating Security-Related Packages

Discovering which packages have security updates available and then updating those packages quickly and easily is important. Yum provides the plug-in for this purpose. The security plug-in extends the yum command with a set of highly-useful security-centric commands, subcommands and options. Refer to Section 6.5.3, “Plug-in Descriptions” for specific information.

6.1.3. Preserving Configuration File Changes

You will inevitably make changes to the configuration files installed by packages as you use your Red Hat Enterprise Linux system. RPM, which Yum uses to perform changes to the system, provides a mechanism for ensuring their integrity. Refer to Section B.2.2, “Installing and Upgrading” for details on how to manage changes to configuration files across package upgrades.

6.1.4. Upgrading the System Off-line with ISO and Yum

For systems that are disconnected from the Internet or Red Hat Network, using the yum update command with the Red Hat Enterprise Linux installation ISO image is an easy and quick way to upgrade systems to the latest minor version. The following steps illustrate the upgrading process:
  1. Create a target directory to mount your ISO image. This directory is not automatically created when mounting, so create it before proceeding to the next step. As root, type:
    mkdir mount_dir
    Replace mount_dir with a path to the mount directory. Typically, users create it as a subdirectory in the /media/ directory.
  2. Mount the Red Hat Enterprise Linux 6 installation ISO image to the previously created target directory. As root, type:
    mount -o loop iso_name mount_dir
    Replace iso_name with a path to your ISO image and mount_dir with a path to the target directory. Here, the -o loop option is required to mount the file as a block device.
  3. Copy the media.repo file from the mount directory to the /etc/yum.repos.d/ directory. Note that configuration files in this directory must have the .repo extension to function properly.
    cp mount_dir/media.repo /etc/yum.repos.d/new.repo
    This creates a configuration file for the yum repository. Replace new.repo with the filename, for example rhel6.repo.
  4. Edit the new configuration file so that it points to the Red Hat Enterprise Linux installation ISO. Add the following line into the /etc/yum.repos.d/new.repo file:
    baseurl=file://mount_dir
    Replace mount_dir with a path to the mount point.
  5. Update all yum repositories including /etc/yum.repos.d/new.repo created in previous steps. As root, type:
    yum update
    This upgrades your system to the version provided by the mounted ISO image.
  6. After successful upgrade, you can unmount the ISO image. As root, type:
    umount mount_dir
    where mount_dir is a path to your mount directory. Also, you can remove the mount directory created in the first step. As root, type:
    rmdir mount_dir
  7. If you will not use the previously created configuration file for another installation or update, you can remove it. As root, type:
    rm /etc/yum.repos.d/new.repo

Example 6.1. Upgrading from Red Hat Enterprise Linux 6.3 to 6.4

Imagine you need to upgrade your system without access to the Internet. To do so, you want to use an ISO image with the newer version of the system, called for instance RHEL6.4-Server-20130130.0-x86_64-DVD1.iso. A target directory created for mounting is /media/rhel6/. As root, change into the directory with your ISO image and type:
~]# mount -o loop RHEL6.4-Server-20130130.0-x86_64-DVD1.iso /media/rhel6/
Then set up a yum repository for your image by copying the media.repo file from the mount directory:
~]# cp /media/rhel6/media.repo /etc/yum.repos.d/rhel6.repo
To make yum recognize the mount point as a repository, add the following line into the /etc/yum.repos.d/rhel6.repo copied in the previous step:
baseurl=file:///media/rhel6/
Now, updating the yum repository will upgrade your system to a version provided by RHEL6.4-Server-20130130.0-x86_64-DVD1.iso. As root, execute:
~]# yum update
When your system is successfully upgraded, you can unmount the image, remove the target directory and the configuration file:
~]# umount /media/rhel6/
~]# rmdir /media/rhel6/
~]# rm /etc/yum.repos.d/rhel6.repo

6.2. Packages and Package Groups

6.2.1. Searching Packages

You can search all RPM package names, descriptions and summaries by using the following command:
yum search term
This command displays the list of matches for each term. For example, to list all packages that match meld or kompare, type:
~]# yum search meld kompare
Loaded plugins: product-id, refresh-packagekit, subscription-manager
Updating Red Hat repositories.
INFO:rhsm-app.repolib:repos updated: 0
============================ Matched: kompare =============================
kdesdk.x86_64 : The KDE Software Development Kit (SDK)
Warning: No matches found for: meld
The yum search command is useful for searching for packages you do not know the name of, but for which you know a related term.

6.2.2. Listing Packages

yum list and related commands provide information about packages, package groups, and repositories.
All of Yum's list commands allow you to filter the results by appending one or more glob expressions as arguments. Glob expressions are normal strings of characters which contain one or more of the wildcard characters * (which expands to match any character multiple times) and ? (which expands to match any one character).

Filtering results with glob expressions

Be careful to escape the glob expressions when passing them as arguments to a yum command, otherwise the Bash shell will interpret these expressions as pathname expansions, and potentially pass all files in the current directory that match the globs to yum. To make sure the glob expressions are passed to yum as intended, either:
  • escape the wildcard characters by preceding them with a backslash character
  • double-quote or single-quote the entire glob expression.
yum list glob_expression
Lists information on installed and available packages matching all glob expressions.

Example 6.2. Listing all ABRT addons and plug-ins using glob expressions

Packages with various ABRT addons and plug-ins either begin with abrt-addon-, or abrt-plugin-. To list these packages, type the following at a shell prompt:
~]# yum list abrt-addon\* abrt-plugin\*
Loaded plugins: product-id, refresh-packagekit, subscription-manager
Updating Red Hat repositories.
INFO:rhsm-app.repolib:repos updated: 0
Installed Packages
abrt-addon-ccpp.x86_64                        1.0.7-5.el6             @rhel
abrt-addon-kerneloops.x86_64                  1.0.7-5.el6             @rhel
abrt-addon-python.x86_64                      1.0.7-5.el6             @rhel
abrt-plugin-bugzilla.x86_64                   1.0.7-5.el6             @rhel
abrt-plugin-logger.x86_64                     1.0.7-5.el6             @rhel
abrt-plugin-sosreport.x86_64                  1.0.7-5.el6             @rhel
abrt-plugin-ticketuploader.x86_64             1.0.7-5.el6             @rhel

yum list all
Lists all installed and available packages.
yum list installed
Lists all packages installed on your system. The rightmost column in the output lists the repository from which the package was retrieved.

Example 6.3. Listing installed packages using a double-quoted glob expression

To list all installed packages that begin with krb followed by exactly one character and a hyphen, type:
~]# yum list installed "krb?-*"
Loaded plugins: product-id, refresh-packagekit, subscription-manager
Updating Red Hat repositories.
INFO:rhsm-app.repolib:repos updated: 0
Installed Packages
krb5-libs.x86_64                         1.8.1-3.el6                  @rhel
krb5-workstation.x86_64                  1.8.1-3.el6                  @rhel

yum list available
Lists all available packages in all enabled repositories.

Example 6.4. Listing available packages using a single glob expression with escaped wildcard characters

To list all available packages with names that contain gstreamer and then plugin, run the following command:
~]# yum list available gstreamer\*plugin\*
Loaded plugins: product-id, refresh-packagekit, subscription-manager
Updating Red Hat repositories.
INFO:rhsm-app.repolib:repos updated: 0
Available Packages
gstreamer-plugins-bad-free.i686               0.10.17-4.el6            rhel
gstreamer-plugins-base.i686                   0.10.26-1.el6            rhel
gstreamer-plugins-base-devel.i686             0.10.26-1.el6            rhel
gstreamer-plugins-base-devel.x86_64           0.10.26-1.el6            rhel
gstreamer-plugins-good.i686                   0.10.18-1.el6            rhel

yum grouplist
Lists all package groups.
yum repolist
Lists the repository ID, name, and number of packages it provides for each enabled repository.

6.2.3. Displaying Package Information

To display information about one or more packages (glob expressions are valid here as well), use the following command:
yum info package_name
For example, to display information about the abrt package, type:
~]# yum info abrt
Loaded plugins: product-id, refresh-packagekit, subscription-manager
Updating Red Hat repositories.
INFO:rhsm-app.repolib:repos updated: 0
Installed Packages
Name       : abrt
Arch       : x86_64
Version    : 1.0.7
Release    : 5.el6
Size       : 578 k
Repo       : installed
From repo  : rhel
Summary    : Automatic bug detection and reporting tool
URL        : https://fedorahosted.org/abrt/
License    : GPLv2+
Description: abrt is a tool to help users to detect defects in applications
           : and to create a bug report with all informations needed by
           : maintainer to fix it. It uses plugin system to extend its
           : functionality.
The yum info package_name command is similar to the rpm -q --info package_name command, but provides as additional information the ID of the Yum repository the RPM package is found in (look for the From repo: line in the output).
You can also query the Yum database for alternative and useful information about a package by using the following command:
yumdb info package_name
This command provides additional information about a package, including the checksum of the package (and algorithm used to produce it, such as SHA-256), the command given on the command line that was invoked to install the package (if any), and the reason that the package is installed on the system (where user indicates it was installed by the user, and dep means it was brought in as a dependency). For example, to display additional information about the yum package, type:
~]# yumdb info yum
Loaded plugins: product-id, refresh-packagekit, subscription-manager
yum-3.2.27-4.el6.noarch
     checksum_data = 23d337ed51a9757bbfbdceb82c4eaca9808ff1009b51e9626d540f44fe95f771
     checksum_type = sha256
     from_repo = rhel
     from_repo_revision = 1298613159
     from_repo_timestamp = 1298614288
     installed_by = 4294967295
     reason = user
     releasever = 6.1
For more information on the yumdb command, refer to the yumdb(8) manual page.

Listing Files Contained in a Package

repoquery is a program for querying information from yum repositories similarly to rpm queries. You can query both package groups and individual packages. To list all files contained in a specific package, type:
repoquery --list package_name
Replace package_name with a name of the package you want to inspect. For more information on the repoquery command, refer to the repoquery manual page.
To find out which package provides a specific file, you can use the yum provides command, described in Finding which package owns a file

6.2.4. Installing Packages

Yum allows you to install both a single package and multiple packages, as well as a package group of your choice.

Installing Individual Packages

To install a single package and all of its non-installed dependencies, enter a command in the following form:
yum install package_name
You can also install multiple packages simultaneously by appending their names as arguments:
yum install package_name package_name
If you are installing packages on a multilib system, such as an AMD64 or Intel64 machine, you can specify the architecture of the package (as long as it is available in an enabled repository) by appending .arch to the package name. For example, to install the sqlite package for i686, type:
~]# yum install sqlite.i686
You can use glob expressions to quickly install multiple similarly-named packages:
~]# yum install perl-Crypt-\*
In addition to package names and glob expressions, you can also provide file names to yum install. If you know the name of the binary you want to install, but not its package name, you can give yum install the path name:
~]# yum install /usr/sbin/named
yum then searches through its package lists, finds the package which provides /usr/sbin/named, if any, and prompts you as to whether you want to install it.

Finding which package owns a file

If you know you want to install the package that contains the named binary, but you do not know in which bin or sbin directory is the file installed, use the yum provides command with a glob expression:
~]# yum provides "*bin/named"
Loaded plugins: product-id, refresh-packagekit, subscription-manager
Updating Red Hat repositories.
INFO:rhsm-app.repolib:repos updated: 0
32:bind-9.7.0-4.P1.el6.x86_64 : The Berkeley Internet Name Domain (BIND)
                              : DNS (Domain Name System) server
Repo        : rhel
Matched from:
Filename    : /usr/sbin/named
yum provides "*/file_name" is a common and useful trick to find the package(s) that contain file_name.

Installing a Package Group

A package group is similar to a package: it is not useful by itself, but installing one pulls a group of dependent packages that serve a common purpose. A package group has a name and a groupid. The yum grouplist -v command lists the names of all package groups, and, next to each of them, their groupid in parentheses. The groupid is always the term in the last pair of parentheses, such as kde-desktop in the following example:
~]# yum -v grouplist kde\*
Loading "product-id" plugin
Loading "refresh-packagekit" plugin
Loading "subscription-manager" plugin
Updating Red Hat repositories.
INFO:rhsm-app.repolib:repos updated: 0
Config time: 0.123
Yum Version: 3.2.29
Setting up Group Process
Looking for repo options for [rhel]
rpmdb time: 0.001
group time: 1.291
Available Groups:
   KDE Desktop (kde-desktop)
Done
You can install a package group by passing its full group name (without the groupid part) to groupinstall:
yum groupinstall group_name
You can also install by groupid:
yum groupinstall groupid
You can even pass the groupid (or quoted name) to the install command if you prepend it with an @-symbol (which tells yum that you want to perform a groupinstall):
yum install @group
For example, the following are alternative but equivalent ways of installing the KDE Desktop group:
~]# yum groupinstall "KDE Desktop"
~]# yum groupinstall kde-desktop
~]# yum install @kde-desktop

6.2.5. Removing Packages

Similarly to package installation, Yum allows you to uninstall (remove in RPM and Yum terminology) both individual packages and a package group.

Removing Individual Packages

To uninstall a particular package, as well as any packages that depend on it, run the following command as root:
yum remove package_name
As when you install multiple packages, you can remove several at once by adding more package names to the command. For example, to remove totem, rhythmbox, and sound-juicer, type the following at a shell prompt:
~]# yum remove totem rhythmbox sound-juicer
Similar to install, remove can take these arguments:
  • package names
  • glob expressions
  • file lists
  • package provides

Removing a package when other packages depend on it

Yum is not able to remove a package without also removing packages which depend on it. This type of operation can only be performed by RPM, is not advised, and can potentially leave your system in a non-functioning state or cause applications to misbehave and/or crash. For further information, refer to Section B.2.4, “Uninstalling” in the RPM chapter.

Removing a Package Group

You can remove a package group using syntax congruent with the install syntax:
yum groupremove group
yum remove @group
The following are alternative but equivalent ways of removing the KDE Desktop group:
~]# yum groupremove "KDE Desktop"
~]# yum groupremove kde-desktop
~]# yum remove @kde-desktop

Intelligent package group removal

When you tell yum to remove a package group, it will remove every package in that group, even if those packages are members of other package groups or dependencies of other installed packages. However, you can instruct yum to remove only those packages which are not required by any other packages or groups by adding the groupremove_leaf_only=1 directive to the [main] section of the /etc/yum.conf configuration file. For more information on this directive, refer to Section 6.4.1, “Setting [main] Options”.

6.3. Working with Transaction History

The yum history command allows users to review information about a timeline of Yum transactions, the dates and times they occurred, the number of packages affected, whether transactions succeeded or were aborted, and if the RPM database was changed between transactions. Additionally, this command can be used to undo or redo certain transactions.

6.3.1. Listing Transactions

To display a list of twenty most recent transactions, as root, either run yum history with no additional arguments, or type the following at a shell prompt:
yum history list
To display all transactions, add the all keyword:
yum history list all
To display only transactions in a given range, use the command in the following form:
yum history list start_id..end_id
You can also list only transactions regarding a particular package or packages. To do so, use the command with a package name or a glob expression:
yum history list glob_expression
For example, the list of the first five transactions looks as follows:
~]# yum history list 1..5
Loaded plugins: product-id, refresh-packagekit, subscription-manager
ID     | Login user               | Date and time    | Action(s)      | Altered
-------------------------------------------------------------------------------
     5 | Jaromir ... <jhradilek>  | 2011-07-29 15:33 | Install        |    1
     4 | Jaromir ... <jhradilek>  | 2011-07-21 15:10 | Install        |    1
     3 | Jaromir ... <jhradilek>  | 2011-07-16 15:27 | I, U           |   73
     2 | System <unset>           | 2011-07-16 15:19 | Update         |    1
     1 | System <unset>           | 2011-07-16 14:38 | Install        | 1106
history list
All forms of the yum history list command produce tabular output with each row consisting of the following columns:
  • ID — an integer value that identifies a particular transaction.
  • Login user — the name of the user whose login session was used to initiate a transaction. This information is typically presented in the Full Name <username> form. For transactions that were not issued by a user (such as an automatic system update), System <unset> is used instead.
  • Date and time — the date and time when a transaction was issued.
  • Action(s) — a list of actions that were performed during a transaction as described in Table 6.1, “Possible values of the Action(s) field”.
  • Altered — the number of packages that were affected by a transaction, possibly followed by additional information as described in Table 6.2, “Possible values of the Altered field”.

Table 6.1. Possible values of the Action(s) field

Action Abbreviation Description
Downgrade D At least one package has been downgraded to an older version.
Erase E At least one package has been removed.
Install I At least one new package has been installed.
Obsoleting O At least one package has been marked as obsolete.
Reinstall R At least one package has been reinstalled.
Update U At least one package has been updated to a newer version.

Table 6.2. Possible values of the Altered field

Symbol Description
< Before the transaction finished, the rpmdb database was changed outside Yum.
> After the transaction finished, the rpmdb database was changed outside Yum.
* The transaction failed to finish.
# The transaction finished successfully, but yum returned a non-zero exit code.
E The transaction finished successfully, but an error or a warning was displayed.
P The transaction finished successfully, but problems already existed in the rpmdb database.
s The transaction finished successfully, but the --skip-broken command line option was used and certain packages were skipped.

Yum also allows you to display a summary of all past transactions. To do so, run the command in the following form as root:
yum history summary
To display only transactions in a given range, type:
yum history summary start_id..end_id
Similarly to the yum history list command, you can also display a summary of transactions regarding a certain package or packages by supplying a package name or a glob expression:
yum history summary glob_expression
For instance, a summary of the transaction history displayed above would look like the following:
~]# yum history summary 1..5
Loaded plugins: product-id, refresh-packagekit, subscription-manager
Login user                 | Time                | Action(s)        | Altered
-------------------------------------------------------------------------------
Jaromir ... <jhradilek>    | Last day            | Install          |        1
Jaromir ... <jhradilek>    | Last week           | Install          |        1
Jaromir ... <jhradilek>    | Last 2 weeks        | I, U             |       73
System <unset>             | Last 2 weeks        | I, U             |     1107
history summary
All forms of the yum history summary command produce simplified tabular output similar to the output of yum history list.
As shown above, both yum history list and yum history summary are oriented towards transactions, and although they allow you to display only transactions related to a given package or packages, they lack important details, such as package versions. To list transactions from the perspective of a package, run the following command as root:
yum history package-list glob_expression
For example, to trace the history of subscription-manager and related packages, type the following at a shell prompt:
~]# yum history package-list subscription-manager\*
Loaded plugins: product-id, refresh-packagekit, subscription-manager
ID     | Action(s)      | Package
-------------------------------------------------------------------------------
     3 | Updated        | subscription-manager-0.95.11-1.el6.x86_64
     3 | Update         |                      0.95.17-1.el6_1.x86_64
     3 | Updated        | subscription-manager-firstboot-0.95.11-1.el6.x86_64
     3 | Update         |                                0.95.17-1.el6_1.x86_64
     3 | Updated        | subscription-manager-gnome-0.95.11-1.el6.x86_64
     3 | Update         |                            0.95.17-1.el6_1.x86_64
     1 | Install        | subscription-manager-0.95.11-1.el6.x86_64
     1 | Install        | subscription-manager-firstboot-0.95.11-1.el6.x86_64
     1 | Install        | subscription-manager-gnome-0.95.11-1.el6.x86_64
history package-list
In this example, three packages were installed during the initial system installation: subscription-manager, subscription-manager-firstboot, and subscription-manager-gnome. In the third transaction, all these packages were updated from version 0.95.11 to version 0.95.17.

6.3.2. Examining Transactions

To display the summary of a single transaction, as root, use the yum history summary command in the following form:
yum history summary id
To examine a particular transaction or transactions in more detail, run the following command as root:
yum history info id
The id argument is optional and when you omit it, yum automatically uses the last transaction. Note that when specifying more than one transaction, you can also use a range:
yum history info start_id..end_id
The following is sample output for two transactions, each installing one new package:
~]# yum history info 4..5
Loaded plugins: product-id, refresh-packagekit, subscription-manager
Transaction ID : 4..5
Begin time     : Thu Jul 21 15:10:46 2011
Begin rpmdb    : 1107:0c67c32219c199f92ed8da7572b4c6df64eacd3a
End time       :            15:33:15 2011 (22 minutes)
End rpmdb      : 1109:1171025bd9b6b5f8db30d063598f590f1c1f3242
User           : Jaromir Hradilek <jhradilek>
Return-Code    : Success
Command Line   : install screen
Command Line   : install yum-plugin-fs-snapshot
Transaction performed with:
    Installed     rpm-4.8.0-16.el6.x86_64
    Installed     yum-3.2.29-17.el6.noarch
    Installed     yum-metadata-parser-1.1.2-16.el6.x86_64
Packages Altered:
    Install screen-4.0.3-16.el6.x86_64
    Install yum-plugin-fs-snapshot-1.1.30-6.el6.noarch
history info
You can also view additional information, such as what configuration options were used at the time of the transaction, or from what repository and why were certain packages installed. To determine what additional information is available for a certain transaction, type the following at a shell prompt as root:
yum history addon-info id
Similarly to yum history info, when no id is provided, yum automatically uses the latest transaction. Another way to refer to the latest transaction is to use the last keyword:
yum history addon-info last
For instance, for the first transaction in the previous example, the yum history addon-info command would provide the following output:
~]# yum history addon-info 4
Loaded plugins: product-id, refresh-packagekit, subscription-manager
Transaction ID: 4
Available additional history information:
  config-main
  config-repos
  saved_tx

history addon-info
In this example, three types of information are available:
  • config-main — global Yum options that were in use during the transaction. Refer to Section 6.4.1, “Setting [main] Options” for information on how to change global options.
  • config-repos — options for individual Yum repositories. Refer to Section 6.4.2, “Setting [repository] Options” for information on how to change options for individual repositories.
  • saved_tx — the data that can be used by the yum load-transaction command in order to repeat the transaction on another machine (see below).
To display selected type of additional information, run the following command as root:
yum history addon-info id information

6.3.3. Reverting and Repeating Transactions

Apart from reviewing the transaction history, the yum history command provides means to revert or repeat a selected transaction. To revert a transaction, type the following at a shell prompt as root:
yum history undo id
To repeat a particular transaction, as root, run the following command:
yum history redo id
Both commands also accept the last keyword to undo or repeat the latest transaction.
Note that both yum history undo and yum history redo commands only revert or repeat the steps that were performed during a transaction. If the transaction installed a new package, the yum history undo command will uninstall it, and if the transaction uninstalled a package the command will again install it. This command also attempts to downgrade all updated packages to their previous version, if these older packages are still available. If you need to restore the system to the state before an update, consider using the fs-snapshot plug-in described in Section 6.5.3, “Plug-in Descriptions”.
When managing several identical systems, Yum also allows you to perform a transaction on one of them, store the transaction details in a file, and after a period of testing, repeat the same transaction on the remaining systems as well. To store the transaction details to a file, type the following at a shell prompt as root:
yum -q history addon-info id saved_tx > file_name
Once you copy this file to the target system, you can repeat the transaction by using the following command as root:
yum load-transaction file_name
Note, however that the rpmdb version stored in the file must be identical to the version on the target system. You can verify the rpmdb version by using the yum version nogroups command.

6.3.4. Completing Transactions

An unexpected situation, such as power loss or system crash, can prevent you from completing your yum transaction. When such event occurs in the middle of your transaction, you can try to resume it later with the following command as root:
yum-complete-transaction
The yum-complete-transaction tool searches for incomplete or aborted yum transactions on a system and attempts to complete them. By default, these transactions are listed in the /var/lib/yum/transaction-all and /var/lib/yum/transaction-done files. If there are more unfinished transactions, yum-complete-transaction attempts to complete the most recent one first.
To clean transaction journal files without attempting to resume the aborted transactions, use the --cleanup-only option:
yum-complete-transaction --cleanup-only

6.3.5. Starting New Transaction History

Yum stores the transaction history in a single SQLite database file. To start new transaction history, run the following command as root:
yum history new
This will create a new, empty database file in the /var/lib/yum/history/ directory. The old transaction history will be kept, but will not be accessible as long as a newer database file is present in the directory.

6.4. Configuring Yum and Yum Repositories

The configuration file for yum and related utilities is located at /etc/yum.conf. This file contains one mandatory [main] section, which allows you to set Yum options that have global effect, and can also contain one or more [repository] sections, which allow you to set repository-specific options. However, it is recommended to define individual repositories in new or existing .repo files in the /etc/yum.repos.d/directory. The values you define in the [main] section of the /etc/yum.conf file can override values set in individual [repository] sections.
This section shows you how to:
  • set global Yum options by editing the [main] section of the /etc/yum.conf configuration file;
  • set options for individual repositories by editing the [repository] sections in /etc/yum.conf and .repo files in the /etc/yum.repos.d/ directory;
  • use Yum variables in /etc/yum.conf and files in the /etc/yum.repos.d/ directory so that dynamic version and architecture values are handled correctly;
  • add, enable, and disable Yum repositories on the command line; and,
  • set up your own custom Yum repository.

6.4.1. Setting [main] Options

The /etc/yum.conf configuration file contains exactly one [main] section, and while some of the key-value pairs in this section affect how yum operates, others affect how Yum treats repositories. You can add many additional options under the [main] section heading in /etc/yum.conf.
A sample /etc/yum.conf configuration file can look like this:
[main]
cachedir=/var/cache/yum/$basearch/$releasever
keepcache=0
debuglevel=2
logfile=/var/log/yum.log
exactarch=1
obsoletes=1
gpgcheck=1
plugins=1
installonly_limit=3

[comments abridged]

# PUT YOUR REPOS HERE OR IN separate files named file.repo
# in /etc/yum.repos.d
The following are the most commonly-used options in the [main] section:
assumeyes=value
…where value is one of:
0yum should prompt for confirmation of critical actions it performs. This is the default.
1 — Do not prompt for confirmation of critical yum actions. If assumeyes=1 is set, yum behaves in the same way that the command line option -y does.
cachedir=directory
…where directory is an absolute path to the directory where Yum should store its cache and database files. By default, Yum's cache directory is /var/cache/yum/$basearch/$releasever.
Refer to Section 6.4.3, “Using Yum Variables” for descriptions of the $basearch and $releasever Yum variables.
debuglevel=value
…where value is an integer between 1 and 10. Setting a higher debuglevel value causes yum to display more detailed debugging output. debuglevel=0 disables debugging output, while debuglevel=2 is the default.
exactarch=value
…where value is one of:
0 — Do not take into account the exact architecture when updating packages.
1 — Consider the exact architecture when updating packages. With this setting, yum will not install an i686 package to update an i386 package already installed on the system. This is the default.
exclude=package_name [more_package_names]
This option allows you to exclude packages by keyword during installation/updates. Listing multiple packages for exclusion can be accomplished by quoting a space-delimited list of packages. Shell globs using wildcards (for example, * and ?) are allowed.
gpgcheck=value
…where value is one of:
0 — Disable GPG signature-checking on packages in all repositories, including local package installation.
1 — Enable GPG signature-checking on all packages in all repositories, including local package installation. gpgcheck=1 is the default, and thus all packages' signatures are checked.
If this option is set in the [main] section of the /etc/yum.conf file, it sets the GPG-checking rule for all repositories. However, you can also set gpgcheck=value for individual repositories instead; that is, you can enable GPG-checking on one repository while disabling it on another. Setting gpgcheck=value for an individual repository in its corresponding .repo file overrides the default if it is present in /etc/yum.conf.
For more information on GPG signature-checking, refer to Section B.3, “Checking a Package's Signature”.
groupremove_leaf_only=value
…where value is one of:
0yum should not check the dependencies of each package when removing a package group. With this setting, yum removes all packages in a package group, regardless of whether those packages are required by other packages or groups. groupremove_leaf_only=0 is the default.
1yum should check the dependencies of each package when removing a package group, and remove only those packages which are not required by any other package or group.
For more information on removing packages, refer to Intelligent package group removal.
installonlypkgs=space separated list of packages
Here you can provide a space-separated list of packages which yum can install, but will never update. Refer to the yum.conf(5) manual page for the list of packages which are install-only by default.
If you add the installonlypkgs directive to /etc/yum.conf, you should ensure that you list all of the packages that should be install-only, including any of those listed under the installonlypkgs section of yum.conf(5). In particular, kernel packages should always be listed in installonlypkgs (as they are by default), and installonly_limit should always be set to a value greater than 2 so that a backup kernel is always available in case the default one fails to boot.
installonly_limit=value
…where value is an integer representing the maximum number of versions that can be installed simultaneously for any single package listed in the installonlypkgs directive.
The defaults for the installonlypkgs directive include several different kernel packages, so be aware that changing the value of installonly_limit will also affect the maximum number of installed versions of any single kernel package. The default value listed in /etc/yum.conf is installonly_limit=3, and it is not recommended to decrease this value, particularly below 2.
keepcache=value
…where value is one of:
0 — Do not retain the cache of headers and packages after a successful installation. This is the default.
1 — Retain the cache after a successful installation.
logfile=file_name
…where file_name is an absolute path to the file in which yum should write its logging output. By default, yum logs to /var/log/yum.log.
multilib_policy=value
…where value is one of:
best — install the best-choice architecture for this system. For example, setting multilib_policy=best on an AMD64 system causes yum to install 64-bit versions of all packages.
all — always install every possible architecture for every package. For example, with multilib_policy set to all on an AMD64 system, yum would install both the i686 and AMD64 versions of a package, if both were available.
obsoletes=value
…where value is one of:
0 — Disable yum's obsoletes processing logic when performing updates.
1 — Enable yum's obsoletes processing logic when performing updates. When one package declares in its spec file that it obsoletes another package, the latter package will be replaced by the former package when the former package is installed. Obsoletes are declared, for example, when a package is renamed. obsoletes=1 the default.
plugins=value
…where value is one of:
0 — Disable all Yum plug-ins globally.

Disabling all plug-ins is not advised

Disabling all plug-ins is not advised because certain plug-ins provide important Yum services. In particular, rhnplugin provides support for RHN Classic, and product-id and subscription-manager plug-ins provide support for the certificate-based Content Delivery Network (CDN). Disabling plug-ins globally is provided as a convenience option, and is generally only recommended when diagnosing a potential problem with Yum.
1 — Enable all Yum plug-ins globally. With plugins=1, you can still disable a specific Yum plug-in by setting enabled=0 in that plug-in's configuration file.
For more information about various Yum plug-ins, refer to Section 6.5, “Yum Plug-ins”. For further information on controlling plug-ins, see Section 6.5.1, “Enabling, Configuring, and Disabling Yum Plug-ins”.
reposdir=directory
…where directory is an absolute path to the directory where .repo files are located. All .repo files contain repository information (similar to the [repository] sections of /etc/yum.conf). yum collects all repository information from .repo files and the [repository] section of the /etc/yum.conf file to create a master list of repositories to use for transactions. If reposdir is not set, yum uses the default directory /etc/yum.repos.d/.
retries=value
…where value is an integer 0 or greater. This value sets the number of times yum should attempt to retrieve a file before returning an error. Setting this to 0 makes yum retry forever. The default value is 10.
For a complete list of available [main] options, refer to the [main] OPTIONS section of the yum.conf(5) manual page.

6.4.2. Setting [repository] Options

The [repository] sections, where repository is a unique repository ID such as my_personal_repo (spaces are not permitted), allow you to define individual Yum repositories.
The following is a bare-minimum example of the form a [repository] section takes:
[repository]
name=repository_name
baseurl=repository_url
Every [repository] section must contain the following directives:
name=repository_name
…where repository_name is a human-readable string describing the repository.
baseurl=repository_url
…where repository_url is a URL to the directory where the repodata directory of a repository is located:
  • If the repository is available over HTTP, use: http://path/to/repo
  • If the repository is available over FTP, use: ftp://path/to/repo
  • If the repository is local to the machine, use: file:///path/to/local/repo
  • If a specific online repository requires basic HTTP authentication, you can specify your username and password by prepending it to the URL as username:password@link. For example, if a repository on http://www.example.com/repo/ requires a username of user and a password of password, then the baseurl link could be specified as http://user:password@www.example.com/repo/.
Usually this URL is an HTTP link, such as:
baseurl=http://path/to/repo/releases/$releasever/server/$basearch/os/
Note that Yum always expands the $releasever, $arch, and $basearch variables in URLs. For more information about Yum variables, refer to Section 6.4.3, “Using Yum Variables”.
Another useful [repository] directive is the following:
enabled=value
…where value is one of:
0 — Do not include this repository as a package source when performing updates and installs. This is an easy way of quickly turning repositories on and off, which is useful when you desire a single package from a repository that you do not want to enable for updates or installs.
1 — Include this repository as a package source.
Turning repositories on and off can also be performed by passing either the --enablerepo=repo_name or --disablerepo=repo_name option to yum, or through the Add/Remove Software window of the PackageKit utility.
Many more [repository] options exist. For a complete list, refer to the [repository] OPTIONS section of the yum.conf(5) manual page.

Example 6.5. A sample /etc/yum.repos.d/redhat.repo file

The following is a sample /etc/yum.repos.d/redhat.repo file:
#
# Red Hat Repositories
# Managed by (rhsm) subscription-manager
#

[red-hat-enterprise-linux-scalable-file-system-for-rhel-6-entitlement-rpms]
name = Red Hat Enterprise Linux Scalable File System (for RHEL 6 Entitlement) (RPMs)
baseurl = https://cdn.redhat.com/content/dist/rhel/entitlement-6/releases/$releasever/$basearch/scalablefilesystem/os
enabled = 1
gpgcheck = 1
gpgkey = file:///etc/pki/rpm-gpg/RPM-GPG-KEY-redhat-release
sslverify = 1
sslcacert = /etc/rhsm/ca/redhat-uep.pem
sslclientkey = /etc/pki/entitlement/key.pem
sslclientcert = /etc/pki/entitlement/11300387955690106.pem

[red-hat-enterprise-linux-scalable-file-system-for-rhel-6-entitlement-source-rpms]
name = Red Hat Enterprise Linux Scalable File System (for RHEL 6 Entitlement) (Source RPMs)
baseurl = https://cdn.redhat.com/content/dist/rhel/entitlement-6/releases/$releasever/$basearch/scalablefilesystem/source/SRPMS
enabled = 0
gpgcheck = 1
gpgkey = file:///etc/pki/rpm-gpg/RPM-GPG-KEY-redhat-release
sslverify = 1
sslcacert = /etc/rhsm/ca/redhat-uep.pem
sslclientkey = /etc/pki/entitlement/key.pem
sslclientcert = /etc/pki/entitlement/11300387955690106.pem

[red-hat-enterprise-linux-scalable-file-system-for-rhel-6-entitlement-debug-rpms]
name = Red Hat Enterprise Linux Scalable File System (for RHEL 6 Entitlement) (Debug RPMs)
baseurl = https://cdn.redhat.com/content/dist/rhel/entitlement-6/releases/$releasever/$basearch/scalablefilesystem/debug
enabled = 0
gpgcheck = 1
gpgkey = file:///etc/pki/rpm-gpg/RPM-GPG-KEY-redhat-release
sslverify = 1
sslcacert = /etc/rhsm/ca/redhat-uep.pem
sslclientkey = /etc/pki/entitlement/key.pem
sslclientcert = /etc/pki/entitlement/11300387955690106.pem

6.4.3. Using Yum Variables

You can use and reference the following built-in variables in yum commands and in all Yum configuration files (that is, /etc/yum.conf and all .repo files in the /etc/yum.repos.d/ directory):
$releasever
You can use this variable to reference the release version of Red Hat Enterprise Linux. Yum obtains the value of $releasever from the distroverpkg=value line in the /etc/yum.conf configuration file. If there is no such line in /etc/yum.conf, then yum infers the correct value by deriving the version number from the redhat-release package.
$arch
You can use this variable to refer to the system's CPU architecture as returned when calling Python's os.uname() function. Valid values for $arch include i686 and x86_64.
$basearch
You can use $basearch to reference the base architecture of the system. For example, i686 machines have a base architecture of i386, and AMD64 and Intel64 machines have a base architecture of x86_64.
$YUM0-9
These ten variables are each replaced with the value of any shell environment variables with the same name. If one of these variables is referenced (in /etc/yum.conf for example) and a shell environment variable with the same name does not exist, then the configuration file variable is not replaced.
To define a custom variable or to override the value of an existing one, create a file with the same name as the variable (without the $ sign) in the /etc/yum/vars/ directory, and add the desired value on its first line.
For example, repository descriptions often include the operating system name. To define a new variable called $osname, create a new file with Red Hat Enterprise Linux on the first line and save it as /etc/yum/vars/osname:
~]# echo "Red Hat Enterprise Linux" > /etc/yum/vars/osname
Instead of Red Hat Enterprise Linux 6, you can now use the following in the .repo files:
name=$osname $releasever

6.4.4. Viewing the Current Configuration

To display the current values of global Yum options (that is, the options specified in the [main] section of the /etc/yum.conf file), run the yum-config-manager with no command line options:
yum-config-manager
To list the content of a different configuration section or sections, use the command in the following form:
yum-config-manager section
You can also use a glob expression to display the configuration of all matching sections:
yum-config-manager glob_expression
For example, to list all configuration options and their corresponding values, type the following at a shell prompt:
~]$ yum-config-manager main \*
Loaded plugins: product-id, refresh-packagekit, subscription-manager
================================== main ===================================
[main]
alwaysprompt = True
assumeyes = False
bandwith = 0
bugtracker_url = https://bugzilla.redhat.com/enter_bug.cgi?product=Red%20Hat%20Enterprise%20Linux%206&component=yum
cache = 0
[output truncated]

6.4.5. Adding, Enabling, and Disabling a Yum Repository

Section 6.4.2, “Setting [repository] Options” described various options you can use to define a Yum repository. This section explains how to add, enable, and disable a repository by using the yum-config-manager command.

The /etc/yum.repos.d/redhat.repo file

When the system is registered with the certificate-based Red Hat Network, the Red Hat Subscription Manager tools are used to manage repositories in the /etc/yum.repos.d/redhat.repo file. Refer to Chapter 5, Registering a System and Managing Subscriptions for more information how to register a system with Red Hat Network and use the Red Hat Subscription Manager tools to manage subscriptions.

Adding a Yum Repository

To define a new repository, you can either add a [repository] section to the /etc/yum.conf file, or to a .repo file in the /etc/yum.repos.d/ directory. All files with the .repo file extension in this directory are read by yum, and it is recommended to define your repositories here instead of in /etc/yum.conf.

Be careful when using untrusted software sources

Obtaining and installing software packages from unverified or untrusted software sources other than Red Hat Network constitutes a potential security risk, and could lead to security, stability, compatibility, and maintainability issues.
Yum repositories commonly provide their own .repo file. To add such a repository to your system and enable it, run the following command as root:
yum-config-manager --add-repo repository_url
…where repository_url is a link to the .repo file. For example, to add a repository located at http://www.example.com/example.repo, type the following at a shell prompt:
~]# yum-config-manager --add-repo http://www.example.com/example.repo
Loaded plugins: product-id, refresh-packagekit, subscription-manager
adding repo from: http://www.example.com/example.repo
grabbing file http://www.example.com/example.repo to /etc/yum.repos.d/example.repo
example.repo                                             |  413 B     00:00
repo saved to /etc/yum.repos.d/example.repo

Enabling a Yum Repository

To enable a particular repository or repositories, type the following at a shell prompt as root:
yum-config-manager --enable repository
…where repository is the unique repository ID (use yum repolist all to list available repository IDs). Alternatively, you can use a glob expression to enable all matching repositories:
yum-config-manager --enable glob_expression
For example, to enable repositories defined in the [example], [example-debuginfo], and [example-source]sections, type:
~]# yum-config-manager --enable example\*
Loaded plugins: product-id, refresh-packagekit, subscription-manager
============================== repo: example ==============================
[example]
bandwidth = 0
base_persistdir = /var/lib/yum/repos/x86_64/6Server
baseurl = http://www.example.com/repo/6Server/x86_64/
cache = 0
cachedir = /var/cache/yum/x86_64/6Server/example
[output truncated]
When successful, the yum-config-manager --enable command displays the current repository configuration.

Disabling a Yum Repository

To disable a Yum repository, run the following command as root:
yum-config-manager --disable repository
…where repository is the unique repository ID (use yum repolist all to list available repository IDs). Similarly to yum-config-manager --enable, you can use a glob expression to disable all matching repositories at the same time:
yum-config-manager --disable glob_expression
When successful, the yum-config-manager --disable command displays the current configuration.

6.4.6. Creating a Yum Repository

To set up a Yum repository, follow these steps:
  1. Install the createrepo package. To do so, type the following at a shell prompt as root:
    yum install createrepo
  2. Copy all packages that you want to have in your repository into one directory, such as /mnt/local_repo/.
  3. Change to this directory and run the following command:
    createrepo --database /mnt/local_repo
    This creates the necessary metadata for your Yum repository, as well as the sqlite database for speeding up yum operations.

    Using the createrepo command on Red Hat Enterprise Linux 5

    Compared to Red Hat Enterprise Linux 5, RPM packages for Red Hat Enterprise Linux 6 are compressed with the XZ lossless data compression format and can be signed with newer hash algorithms like SHA-256. Consequently, it is not recommended to use the createrepo command on Red Hat Enterprise Linux 5 to create the package metadata for Red Hat Enterprise Linux 6. If you want to use createrepo on this system anyway, install the python-hashlib package from EPEL (Extra Packages for Enterprise Linux) so that the repodata can also use the SHA-256 hash algorithm.

6.4.7. Working with Yum Cache

By default, yum deletes downloaded data files when they are no longer needed after a successful operation. This minimizes the amount of storage space that yum uses. However, you can enable caching, so that the package files downloaded by yum stay in cache directories. By using cached data, you can carry out certain operations without a network connection, you can also copy packages stored in the caches and reuse them elsewhere
Yum stores temporary files in the /var/cache/yum/$basearch/$releasever/ directory, where $basearch and $releasever are Yum variables refering to base architecture of the system and the release version of Red Hat Enterprise Linux. Each configured repository has one subdirectory. For example, the directory /var/cache/yum/$basearch/$releasever/development/packages/ holds packages downloaded from the development repository. You can find the values for the $basearch and $releasever variables in the output of the yum version command.
To change the default cache location, modify the cachedir option in the [main] section of the /etc/yum.conf configuration file. Refer to Section 6.4, “Configuring Yum and Yum Repositories” for more information on configuring yum.

Enabling the Caches

To retain the cache of packages after a successful installation, add the following text to the [main] section of /etc/yum.conf.
keepcache = 1
Once you enabled caching, every yum operation may download package data from the configured repositories.
To download and make usable all the metadata for the currently enabled yum repositories, type:
yum makecache
This is useful if you want to make sure that the cache is fully up to date with all metadata. To set the time after which the metadata will expire, use the metadata-expire setting in /etc/yum.conf.

Using yum in Cache-only Mode

To carry out a yum command without a network connection, add the -C or --cacheonly command-line option. With this option, yum proceeds without checking any network repositories, and uses only cached files. In this mode, yum may only install packages that have been downloaded and cached by a previous operation.
For instance, to list packages that use the currently cached data with names that contain “gstreamer”, enter the following command:
yum -C list gstreamer*

Clearing the yum Caches

It is often useful to remove entries accumulated in the /var/cache/yum/ directory. If you remove a package from the cache, you do not affect the copy of the software installed on your system. To remove all entries for currently enabled repositories from the cache, type the following as a root:
yum clean all
There are various ways to invoke yum in clean mode depending on the type of cached data you want to remove. Refer to Table 6.3, “Available yum clean options” for a complete list of available configuration options.

Table 6.3. Available yum clean options

Option Description
expire-cache eliminates time records of the metadata and mirrorlists download for each repository. This forces yum to revalidate the cache for each repository the next time it is used.
packages eliminates any cached packages from the system
headers eliminates all header files that previous versions of yum used for dependency resolution
metadata eliminates all files that yum uses to determine the remote availability of packages. These metadata are downloaded again the next time yum is run.
dbcache eliminates the sqlite cache used for faster access to metadata. Using this option will force yum to download the sqlite metadata the next time it is run. This does not apply for repositories that contain only .xml data, in that case, sqlite data are deleted but without subsequent download
rpmdb eliminates any cached data from the local rpmdb
plugins enabled plugins are forced to eliminate their cached data
all removes all of the above

The expire-cache option is most preferred from the above list. In many cases, it is a sufficient and much faster replacement for clean all.

6.5. Yum Plug-ins

Yum provides plug-ins that extend and enhance its operations. Certain plug-ins are installed by default. Yum always informs you which plug-ins, if any, are loaded and active whenever you call any yum command. For example:
~]# yum info yum
Loaded plugins: product-id, refresh-packagekit, subscription-manager
[output truncated]
Note that the plug-in names which follow Loaded plugins are the names you can provide to the --disableplugins=plugin_name option.

6.5.1. Enabling, Configuring, and Disabling Yum Plug-ins

To enable Yum plug-ins, ensure that a line beginning with plugins= is present in the [main] section of /etc/yum.conf, and that its value is 1:
plugins=1
You can disable all plug-ins by changing this line to plugins=0.

Disabling all plug-ins is not advised

Disabling all plug-ins is not advised because certain plug-ins provide important Yum services. In particular, rhnplugin provides support for RHN Classic, and product-id and subscription-manager plug-ins provide support for the certificate-based Content Delivery Network (CDN). Disabling plug-ins globally is provided as a convenience option, and is generally only recommended when diagnosing a potential problem with Yum.
Every installed plug-in has its own configuration file in the /etc/yum/pluginconf.d/ directory. You can set plug-in specific options in these files. For example, here is the refresh-packagekit plug-in's refresh-packagekit.conf configuration file:
[main]
enabled=1
Plug-in configuration files always contain a [main] section (similar to Yum's /etc/yum.conf file) in which there is (or you can place if it is missing) an enabled= option that controls whether the plug-in is enabled when you run yum commands.
If you disable all plug-ins by setting enabled=0 in /etc/yum.conf, then all plug-ins are disabled regardless of whether they are enabled in their individual configuration files.
If you merely want to disable all Yum plug-ins for a single yum command, use the --noplugins option.
If you want to disable one or more Yum plug-ins for a single yum command, add the --disableplugin=plugin_name option to the command. For example, to disable the presto plug-in while updating a system, type:
~]# yum update --disableplugin=presto
The plug-in names you provide to the --disableplugin= option are the same names listed after the Loaded plugins line in the output of any yum command. You can disable multiple plug-ins by separating their names with commas. In addition, you can match multiple plug-in names or shorten long ones by using glob expressions:
~]# yum update --disableplugin=presto,refresh-pack*

6.5.2. Installing Additional Yum Plug-ins

Yum plug-ins usually adhere to the yum-plugin-plugin_name package-naming convention, but not always: the package which provides the presto plug-in is named yum-presto, for example. You can install a Yum plug-in in the same way you install other packages. For instance, to install the security plug-in, type the following at a shell prompt:
~]# yum install yum-plugin-security

6.5.3. Plug-in Descriptions

The following list provides descriptions of a few useful Yum plug-ins:
fs-snapshot (yum-plugin-fs-snapshot)
The fs-snapshot plug-in extends Yum to create a snapshot of a file system before proceeding with a transaction such as a system update or package removal. When a user decides that the changes made by the transaction are unwanted, this mechanism allows the user to roll back to the changes that are stored in a snapshot.
In order for the plug-in to work, the root file system (that is, /) must be on an LVM (Logical Volume Manager) or Btrfs volume. To use the fs-snapshot plug-in on an LVM volume, take the following steps:
  1. Make sure that the volume group with the root file system has enough free extents. The required size is a function of the amount of changes to the original logical volume that is expected during the life of the snapshot. The reasonable default is 50–80 % of the original logical volume size.
    To display detailed information about a particular volume group, run the vgdisplay command in the following form as root:
    vgdisplay volume_group
    The number of free extents is listed on the Free PE / Size line.
  2. If the volume group with the root file system does not have enough free extents, add a new physical volume:
    1. As root, run the pvcreate command in the following form to initialize a physical volume for use with the Logical Volume Manager:
      pvcreate device
    2. Use the vgextend command in the following form as root to add the physical volume to the volume group:
      vgextend volume_group physical_volume
  3. Edit the configuration file located in /etc/yum/pluginconf.d/fs-snapshot.conf, and make the following changes to the [lvm] section:
    1. Change the value of the enabled option to 1:
      enabled = 1
    2. Remove the hash sign (that is, #) from the beginning of the lvcreate_size_args line, and adjust the number of logical extents which are allocated for a snapshot. For example, to allocate 80 % of the size of the original logical volume, use:
      lvcreate_size_args = -l 80%ORIGIN
    Refer to Table 6.4, “Supported fs-snapshot.conf directives” for a complete list of available configuration options.
  4. Run the desired yum command, and make sure fs-snapshot is included in the list of loaded plug-ins (the Loaded plugins line) before you confirm the changes and proceed with the transaction. The fs-snapshot plug-in displays a line in the following form for each affected logical volume:
    fs-snapshot: snapshotting file_system (/dev/volume_group/logical_volume): logical_volume_yum_timestamp
  5. Verify that the system is working as expected:
    • If you decide to keep the changes, remove the snapshot by running the lvremove command as root:
      lvremove /dev/volume_group/logical_volume_yum_timestamp
    • If you decide to revert the changes and restore the file system to a state that is saved in a snapshot, take the following steps:
      1. As root, run the command in the following form to merge a snapshot into its original logical volume:
        lvconvert --merge /dev/volume_group/logical_volume_yum_timestamp
        The lvconvert command will inform you that a restart is required in order for the changes to take effect.
      2. Restart the system as instructed. You can do so by typing the following at a shell prompt as root:
        reboot
To use the fs-snapshot plug-in on a Btrfs file system, take the following steps:
  1. Run the desired yum command, and make sure fs-snapshot is included in the list of loaded plug-ins (the Loaded plugins line) before you confirm the changes and proceed with the transaction. The fs-snapshot plug-in displays a line in the following form for each affected file system:
    fs-snapshot: snapshotting file_system: file_system/yum_timestamp
  2. Verify that the system is working as expected:
    • If you decide to keep the changes, you can optionally remove unwanted snapshots. To remove a Btrfs snapshot, use the command in the following form as root:
      btrfs subvolume delete file_system/yum_timestamp
    • If you decide to revert the changes and restore a file system to a state that is saved in a snapshot, take the following steps:
      1. Determine the identifier of a particular snapshot by using the following command as root:
        btrfs subvolume list file_system
      2. As root, configure the system to mount this snapshot by default:
        btrfs subvolume set-default id file_system
      3. Restart the system. You can do so by typing the following at a shell prompt as root:
        reboot
Note that in Red Hat Enterprise Linux 6, Btrfs is included as a technology preview to allow you to experiment with this file system, and is only available on 64-bit x86 architectures. Do not use Btrfs for partitions that will contain valuable data or that are essential for the operation of important systems.
For more information on logical volume management, Btrfs, and file system snapshots, see the Red Hat Enterprise Linux 6 Storage Administration Guide. For additional information about the plug-in and its configuration, refer to the yum-fs-snapshot(1) and yum-fs-snapshot.conf(5) manual pages.

Table 6.4. Supported fs-snapshot.conf directives

Section Directive Description
[main] enabled=value Allows you to enable or disable the plug-in. The value must be either 1 (enabled), or 0 (disabled). When installed, the plug-in is enabled by default.
exclude=list Allows you to exclude certain file systems. The value must be a space-separated list of mount points you do not want to snapshot (for example, /srv /mnt/backup). This option is not included in the configuration file by default.
[lvm] enabled=value Allows you to enable or disable the use of the plug-in on LVM volumes. The value must be either 1 (enabled), or 0 (disabled). This option is disabled by default.
lvcreate_size_args=value Allows you to specify the size of a logical volume snapshot. The value must be the -l or -L command line option for the lvcreate utility followed by a valid argument (for example, -l 80%ORIGIN).

kabi (kabi-yum-plugins)
The kabi plug-in checks whether a driver update package conforms with official Red Hat kernel Application Binary Interface (kABI). With this plug-in enabled, when a user attempts to install a package that uses kernel symbols which are not on a whitelist, a warning message is written to the system log. Additionally, configuring the plug-in to run in enforcing mode prevents such packages from being installed at all.
To configure the kabi plug-in, edit the configuration file located in /etc/yum/pluginconf.d/kabi.conf. Refer to Table 6.5, “Supported kabi.conf directives” for a list of directives that can be used in the [main] section.

Table 6.5. Supported kabi.conf directives

Directive Description
enabled=value Allows you to enable or disable the plug-in. The value must be either 1 (enabled), or 0 (disabled). When installed, the plug-in is enabled by default.
whitelists=directory Allows you to specify the directory in which the files with supported kernel symbols are located. By default, the kabi plug-in uses files provided by the kabi-whitelists package (that is, the /lib/modules/kabi/ directory).
enforce=value Allows you to enable or disable enforcing mode. The value must be either 1 (enabled), or 0 (disabled). By default, this option is commented out and the kabi plug-in only displays a warning message.

presto (yum-presto)
The presto plug-in adds support to Yum for downloading delta RPM packages, during updates, from repositories which have presto metadata enabled. Delta RPMs contain only the differences between the version of the package installed on the client requesting the RPM package and the updated version in the repository.
Downloading a delta RPM is much quicker than downloading the entire updated package, and can speed up updates considerably. Once the delta RPMs are downloaded, they must be rebuilt to apply the difference to the currently-installed package and thus create the full, updated package. This process takes CPU time on the installing machine. Using delta RPMs is therefore a tradeoff between time-to-download, which depends on the network connection, and time-to-rebuild, which is CPU-bound. Using the presto plug-in is recommended for fast machines and systems with slower network connections, while slower machines on very fast connections benefit more from downloading normal RPM packages, that is, by disabling presto.
product-id (subscription-manager)
The product-id plug-in manages product identity certificates for products installed from the Content Delivery Network. The product-id plug-in is installed by default.
protect-packages (yum-plugin-protect-packages)
The protect-packages plug-in prevents the yum package and all packages it depends on from being deliberately or accidentally removed. This prevents many of the most important packages necessary for your system to run from being removed. In addition, you can list more packages, one per line, in the /etc/sysconfig/protected-packages file[2] (which you should create if it does not exist), and protect-packages will extend protection-from-removal to those packages as well.
To temporarily override package protection, use the --override-protection option with an applicable yum command.
refresh-packagekit (PackageKit-yum-plugin)
The refresh-packagekit plug-in updates metadata for PackageKit whenever yum is run. The refresh-packagekit plug-in is installed by default.
rhnplugin (yum-rhn-plugin)
The rhnplugin provides support for connecting to RHN Classic. This allows systems registered with RHN Classic to update and install packages from this system. Note that RHN Classic is only provided for older Red Hat Enterprise Linux systems (that is, Red Hat Enterprise Linux 4.x, Red Hat Enterprise Linux 5.x, and Satellite 5.x) in order to migrate them over to Red Hat Enterprise Linux 6. The rhnplugin is installed by default.
Refer to the rhnplugin(8) manual page for more information about the plug-in.
security (yum-plugin-security)
Discovering information about and applying security updates easily and often is important to all system administrators. For this reason Yum provides the security plug-in, which extends yum with a set of highly-useful security-related commands, subcommands and options.
You can check for security-related updates as follows:
~]# yum check-update --security
Loaded plugins: product-id, refresh-packagekit, security, subscription-manager
Updating Red Hat repositories.
INFO:rhsm-app.repolib:repos updated: 0
Limiting package lists to security relevant ones
Needed 3 of 7 packages, for security
elinks.x86_64                   0.12-0.13.el6               rhel
kernel.x86_64                   2.6.30.8-64.el6             rhel
kernel-headers.x86_64           2.6.30.8-64.el6             rhel
You can then use either yum update --security or yum update-minimal --security to update those packages which are affected by security advisories. Both of these commands update all packages on the system for which a security advisory has been issued. yum update-minimal --security updates them to the latest packages which were released as part of a security advisory, while yum update --security will update all packages affected by a security advisory to the latest version of that package available.
In other words, if:
  • the kernel-2.6.30.8-16 package is installed on your system;
  • the kernel-2.6.30.8-32 package was released as a security update;
  • then kernel-2.6.30.8-64 was released as a bug fix update,
...then yum update-minimal --security will update you to kernel-2.6.30.8-32, and yum update --security will update you to kernel-2.6.30.8-64. Conservative system administrators probably want to use update-minimal to reduce the risk incurred by updating packages as much as possible.
Refer to the yum-security(8) manual page for usage details and further explanation of the enhancements the security plug-in adds to yum.
subscription-manager (subscription-manager)
The subscription-manager plug-in provides support for connecting to Red Hat Network. This allows systems registered with Red Hat Network to update and install packages from the certificate-based Content Delivery Network. The subscription-manager plug-in is installed by default.
Refer to Chapter 5, Registering a System and Managing Subscriptions for more information how to manage product subscriptions and entitlements.
yum-downloadonly (yum-plugin-downloadonly)
The yum-downloadonly plug-in provides the --downloadonly command-line option which can be used to download packages from Red Hat Network or a configured Yum repository without installing the packages.
To install the package, follow the instructions in Section 6.5.2, “Installing Additional Yum Plug-ins”. After the installation, see the contents of the /etc/yum/pluginconf.d/downloadonly.conf file to ensure that the plug-in is enabled:
~]$ cat /etc/yum/pluginconf.d/downloadonly.conf
[main]
enabled=1
In the following example, the yum install --downloadonly command is run to download the latest version of the httpd package, without installing it:
~]# yum install httpd --downloadonly
Loaded plugins: downloadonly, product-id, refresh-packagekit, rhnplugin,
              : subscription-manager
Updating Red Hat repositories.
Setting up Install Process
Resolving Dependencies
--> Running transaction check
---> Package httpd.x86_64 0:2.2.15-9.el6_1.2 will be updated
---> Package httpd.x86_64 0:2.2.15-15.el6_2.1 will be an update
--> Processing Dependency: httpd-tools = 2.2.15-15.el6_2.1 for package: httpd-2.2.15-15.el6_2.1.x86_64
--> Running transaction check
---> Package httpd-tools.x86_64 0:2.2.15-9.el6_1.2 will be updated
---> Package httpd-tools.x86_64 0:2.2.15-15.el6_2.1 will be an update
--> Finished Dependency Resolution

Dependencies Resolved

================================================================================
 Package        Arch      Version                 Repository               Size
================================================================================
Updating:
 httpd          x86_64    2.2.15-15.el6_2.1       rhel-x86_64-server-6    812 k
Updating for dependencies:
 httpd-tools    x86_64    2.2.15-15.el6_2.1       rhel-x86_64-server-6     70 k

Transaction Summary
================================================================================
Upgrade       2 Package(s)

Total download size: 882 k
Is this ok [y/N]: y
Downloading Packages:
(1/2): httpd-2.2.15-15.el6_2.1.x86_64.rpm                | 812 kB     00:00
(2/2): httpd-tools-2.2.15-15.el6_2.1.x86_64.rpm          |  70 kB     00:00
--------------------------------------------------------------------------------
Total                                           301 kB/s | 882 kB     00:02


exiting because --downloadonly specified
By default, packages downloaded using the --downloadonly option are saved in one of the subdirectories of the /var/cache/yum directory, depending on the Red Hat Enterprise Linux variant and architecture.
If you want to specify an alternate directory to save the packages, pass the --downloaddir option allong with --downloadonly:
~]# yum install --donwloadonly --downloaddir=/path/to/direcotry httpd

Note

As an alternative to the yum-downloadonly plugin — to download packages without installing them — you can use the yumdownloader utility that is provided by the yum-utils package.

6.6. Additional Resources

http://yum.baseurl.org/wiki/Guides
The Yum Guides section of the Yum wiki contains more documentation.


[2] You can also place files with the extension .list in the /etc/sysconfig/protected-packages.d/ directory (which you should create if it does not exist), and list packages—one per line—in these files. protect-packages will protect these too.

Chapter 7. PackageKit

Red Hat provides PackageKit for viewing, managing, updating, installing and uninstalling packages compatible with your system. PackageKit consists of several graphical interfaces that can be opened from the GNOME panel menu, or from the Notification Area when PackageKit alerts you that updates are available. For more information on PackageKit's architecture and available front ends, refer to Section 7.3, “PackageKit Architecture”.

7.1. Updating Packages with Software Update

PackageKit displays a starburst icon in the Notification Area whenever updates are available to be installed on your system.
PackageKit's icon in the Notification Area

Figure 7.1. PackageKit's icon in the Notification Area


Clicking on the notification icon opens the Software Update window. Alternatively, you can open Software Updates by clicking SystemAdministrationSoftware Update from the GNOME panel, or running the gpk-update-viewer command at the shell prompt. In the Software Updates window, all available updates are listed along with the names of the packages being updated (minus the .rpm suffix, but including the CPU architecture), a short summary of the package, and, usually, short descriptions of the changes the update provides. Any updates you do not wish to install can be de-selected here by unchecking the checkbox corresponding to the update.
Installing updates with Software Update

Figure 7.2. Installing updates with Software Update


The updates presented in the Software Updates window only represent the currently-installed packages on your system for which updates are available; dependencies of those packages, whether they are existing packages on your system or new ones, are not shown until you click Install Updates.
PackageKit utilizes the fine-grained user authentication capabilities provided by the PolicyKit toolkit whenever you request it to make changes to the system. Whenever you instruct PackageKit to update, install or remove packages, you will be prompted to enter the superuser password before changes are made to the system.
If you instruct PackageKit to update the kernel package, then it will prompt you after installation, asking you whether you want to reboot the system and thereby boot into the newly-installed kernel.

Setting the Update-Checking Interval

Right-clicking on PackageKit's Notification Area icon and clicking Preferences opens the Software Update Preferences window, where you can define the interval at which PackageKit checks for package updates, as well as whether or not to automatically install all updates or only security updates. Leaving the Check for updates when using mobile broadband box unchecked is handy for avoiding extraneous bandwidth usage when using a wireless connection on which you are charged for the amount of data you download.
Setting PackageKit's update-checking interval

Figure 7.3. Setting PackageKit's update-checking interval


7.2. Using Add/Remove Software

To find and install a new package, on the GNOME panel click on SystemAdministrationAdd/Remove Software, or run the gpk-application command at the shell prompt.
PackageKit's Add/Remove Software window

Figure 7.4. PackageKit's Add/Remove Software window


7.2.1. Refreshing Software Sources (Yum Repositories)

PackageKit refers to Yum repositories as software sources. It obtains all packages from enabled software sources. You can view the list of all configured and unfiltered (see below) Yum repositories by opening Add/Remove Software and clicking SystemSoftware sources. The Software Sources dialog shows the repository name, as written on the name=<My Repository Name> field of all [repository] sections in the /etc/yum.conf configuration file, and in all repository.repo files in the /etc/yum.repos.d/ directory.
Entries which are checked in the Enabled column indicate that the corresponding repository will be used to locate packages to satisfy all update and installation requests (including dependency resolution). You can enable or disable any of the listed Yum repositories by selecting or clearing the checkbox. Note that doing so causes PolicyKit to prompt you for superuser authentication.
The Enabled column corresponds to the enabled=<1 or 0> field in [repository] sections. When you click the checkbox, PackageKit inserts the enabled=<1 or 0> line into the correct [repository] section if it does not exist, or changes the value if it does. This means that enabling or disabling a repository through the Software Sources window causes that change to persist after closing the window or rebooting the system.
Note that it is not possible to add or remove Yum repositories through PackageKit.

Showing source RPM, test and debuginfo repositories

Checking the box at the bottom of the Software Sources window causes PackageKit to display source RPM, testing and debuginfo repositories as well. This box is unchecked by default.
After making a change to the available Yum repositories, click on SystemRefresh package lists to make sure your package list is up-to-date.

7.2.2. Finding Packages with Filters

Once the software sources have been updated, it is often beneficial to apply some filters so that PackageKit retrieves the results of our Find queries faster. This is especially helpful when performing many package searches. Four of the filters in the Filters drop-down menu are used to split results by matching or not matching a single criterion. By default when PackageKit starts, these filters are all unapplied (No filter), but once you do filter by one of them, that filter remains set until you either change it or close PackageKit.
Because you are usually searching for available packages that are not installed on the system, click FiltersInstalled and select the Only available radio button.
Filtering out already-installed packages

Figure 7.5. Filtering out already-installed packages


Also, unless you require development files such as C header files, click FiltersDevelopment and select the Only end user files radio button. This filters out all of the <package_name>-devel packages we are not interested in.
Filtering out development packages from the list of Find results

Figure 7.6. Filtering out development packages from the list of Find results


The two remaining filters with submenus are:
Graphical
Narrows the search to either applications which provide a GUI interface (Only graphical) or those that do not. This filter is useful when browsing for GUI applications that perform a specific function.
Free
Search for packages which are considered to be free software. Refer to the Fedora Licensing List for details on approved licenses.
The remaining filters can be enabled by selecting the checkboxes next to them:
Hide subpackages
Checking the Hide subpackages checkbox filters out generally-uninteresting packages that are typically only dependencies of other packages that we want. For example, checking Hide subpackages and searching for <package> would cause the following related packages to be filtered out of the Find results (if it exists):
  • <package>-devel
  • <package>-libs
  • <package>-libs-devel
  • <package>-debuginfo
Only newest packages
Checking Only newest packages filters out all older versions of the same package from the list of results, which is generally what we want. Note that this filter is often combined with the Only available filter to search for the latest available versions of new (not installed) packages.
Only native packages
Checking the Only native packages box on a multilib system causes PackageKit to omit listing results for packages compiled for the architecture that runs in compatibility mode. For example, enabling this filter on a 64-bit system with an AMD64 CPU would cause all packages built for the 32-bit x86 CPU architecture not to be shown in the list of results, even though those packages are able to run on an AMD64 machine. Packages which are architecture-agnostic (i.e. noarch packages such as crontabs-1.10-32.1.el6.noarch.rpm) are never filtered out by checking Only native packages. This filter has no affect on non-multilib systems, such as x86 machines.

7.2.3. Installing and Removing Packages (and Dependencies)

With the two filters selected, Only available and Only end user files, search for the screen window manager for the command line and highlight the package. You now have access to some very useful information about it, including: a clickable link to the project homepage; the Yum package group it is found in, if any; the license of the package; a pointer to the GNOME menu location from where the application can be opened, if applicable; and the size of the package, which is relevant when we download and install it.
Viewing and installing a package with PackageKit's Add/Remove Software window

Figure 7.7. Viewing and installing a package with PackageKit's Add/Remove Software window


When the checkbox next to a package or group is checked, then that item is already installed on the system. Checking an unchecked box causes it to be marked for installation, which only occurs when the Apply button is clicked. In this way, you can search for and select multiple packages or package groups before performing the actual installation transactions. Additionally, you can remove installed packages by unchecking the checked box, and the removal will occur along with any pending installations when Apply is pressed. Dependency resolution, which may add additional packages to be installed or removed, is performed after pressing Apply. PackageKit will then display a window listing those additional packages to install or remove, and ask for confirmation to proceed.
Select screen and click the Apply button. You will then be prompted for the superuser password; enter it, and PackageKit will install screen. After finishing the installation, PackageKit sometimes presents you with a list of your newly-installed applications and offers you the choice of running them immediately. Alternatively, you will remember that finding a package and selecting it in the Add/Remove Software window shows you the Location of where in the GNOME menus its application shortcut is located, which is helpful when you want to run it.
Once it is installed, you can run screen, a screen manager that allows you to have multiple logins on one terminal, by typing screen at a shell prompt.
screen is a very useful utility, but we decide that we do not need it and we want to uninstall it. Remembering that we need to change the Only available filter we recently used to install it to Only installed in FiltersInstalled, we search for screen again and uncheck it. The program did not install any dependencies of its own; if it had, those would be automatically removed as well, as long as they were not also dependencies of any other packages still installed on our system.

Removing a package when other packages depend on it

Although PackageKit automatically resolves dependencies during package installation and removal, it is unable to remove a package without also removing packages which depend on it. This type of operation can only be performed by RPM, is not advised, and can potentially leave your system in a non-functioning state or cause applications to behave erratically and/or crash.
Removing a package with PackageKit's Add/Remove Software window

Figure 7.8. Removing a package with PackageKit's Add/Remove Software window


7.2.4. Installing and Removing Package Groups

PackageKit also has the ability to install Yum package groups, which it calls Package collections. Clicking on Package collections in the top-left list of categories in the Software Updates window allows us to scroll through and find the package group we want to install. In this case, we want to install Czech language support (the Czech Support group). Checking the box and clicking apply informs us how many additional packages must be installed in order to fulfill the dependencies of the package group.
Installing the Czech Support package group

Figure 7.9. Installing the Czech Support package group


Similarly, installed package groups can be uninstalled by selecting Package collections, unchecking the appropriate checkbox, and applying.

7.2.5. Viewing the Transaction Log

PackageKit maintains a log of the transactions that it performs. To view the log, from the Add/Remove Software window, click SystemSoftware log, or run the gpk-log command at the shell prompt.
The Software Log Viewer shows the following information:
  • Date — the date on which the transaction was performed.
  • Action — the action that was performed during the transaction, for example Updated packages or Installed packages.
  • Details — the transaction type such as Updated, Installed, or Removed, followed by a list of affected packages.
  • Username — the name of the user who performed the action.
  • Application — the front end application that was used to perform the action, for example Update System.
Typing the name of a package in the top text entry field filters the list of transactions to those which affected that package.
Viewing the log of package management transactions with the Software Log Viewer

Figure 7.10. Viewing the log of package management transactions with the Software Log Viewer


7.3. PackageKit Architecture

Red Hat provides the PackageKit suite of applications for viewing, updating, installing and uninstalling packages and package groups compatible with your system. Architecturally, PackageKit consists of several graphical front ends that communicate with the packagekitd daemon back end, which communicates with a package manager-specific back end that utilizes Yum to perform the actual transactions, such as installing and removing packages, etc.
Table 7.1, “PackageKit GUI windows, menu locations, and shell prompt commands” shows the name of the GUI window, how to start the window from the GNOME desktop or from the Add/Remove Software window, and the name of the command line application that opens that window.

Table 7.1. PackageKit GUI windows, menu locations, and shell prompt commands

Window Title Function How to Open Shell Command
Add/Remove Software Install, remove or view package info
From the GNOME panel: SystemAdministrationAdd/Remove Software
gpk-application
Software Update Perform package updates
From the GNOME panel: SystemAdministrationSoftware Update
gpk-update-viewer
Software Sources Enable and disable Yum repositories
From Add/Remove Software: SystemSoftware Sources
gpk-repo
Software Log Viewer View the transaction log
From Add/Remove Software: SystemSoftware Log
gpk-log
Software Update Preferences Set PackageKit preferences gpk-prefs
(Notification Area Alert) Alerts you when updates are available
From the GNOME panel: SystemPreferencesStartup Applications, the Startup Programs tab
gpk-update-icon

The packagekitd daemon runs outside the user session and communicates with the various graphical front ends. The packagekitd daemon[3] communicates via the DBus system message bus with another back end, which utilizes Yum's Python API to perform queries and make changes to the system. On Linux systems other than Red Hat and Fedora, packagekitd can communicate with other back ends that are able to utilize the native package manager for that system. This modular architecture provides the abstraction necessary for the graphical interfaces to work with many different package managers to perform essentially the same types of package management tasks. Learning how to use the PackageKit front ends means that you can use the same familiar graphical interface across many different Linux distributions, even when they utilize a native package manager other than Yum.
In addition, PackageKit's separation of concerns provides reliability in that a crash of one of the GUI windows—or even the user's X Window session—will not affect any package management tasks being supervised by the packagekitd daemon, which runs outside of the user session.
All of the front end graphical applications discussed in this chapter are provided by the gnome-packagekit package instead of by PackageKit and its dependencies.
Finally, PackageKit also comes with a console-based front end called pkcon.

7.4. Additional Resources

PackageKit home page — http://www.packagekit.org/index.html
Information about and mailing lists for PackageKit.
PackageKit FAQ — http://www.packagekit.org/pk-faq.html
An informative list of Frequently Asked Questions for the PackageKit software suite.
PackageKit Feature Matrix — http://www.packagekit.org/pk-matrix.html
Cross-reference PackageKit-provided features with the long list of package manager back ends.


[3] System daemons are typically long-running processes that provide services to the user or to other programs, and which are started, often at boot time, by special initialization scripts (often shortened to init scripts). Daemons respond to the service command and can be turned on or off permanently by using the chkconfig on or chkconfig offcommands. They can typically be recognized by a d appended to their name, such as the packagekitd daemon. Refer to Chapter 11, Services and Daemons for information about system services.

Part III. Networking

This part describes how to configure the network on Red Hat Enterprise Linux.

Table of Contents

8. NetworkManager
8.1. The NetworkManager Daemon
8.2. Interacting with NetworkManager
8.2.1. Connecting to a Network
8.2.2. Configuring New and Editing Existing Connections
8.2.3. Connecting to a Network Automatically
8.2.4. User and System Connections
8.3. Establishing Connections
8.3.1. Establishing a Wired (Ethernet) Connection
8.3.2. Establishing a Wireless Connection
8.3.3. Establishing a Mobile Broadband Connection
8.3.4. Establishing a VPN Connection
8.3.5. Establishing a DSL Connection
8.3.6. Establishing a Bond Connection
8.3.7. Establishing a VLAN Connection
8.3.8. Establishing an IP-over-InfiniBand (IPoIB) Connection
8.3.9. Configuring Connection Settings
8.4. NetworkManager Architecture
9. Network Interfaces
9.1. Network Configuration Files
9.2. Interface Configuration Files
9.2.1. Ethernet Interfaces
9.2.2. Specific ifcfg Options for Linux on System z
9.2.3. Required ifcfg Options for Linux on System z
9.2.4. Channel Bonding Interfaces
9.2.5. Network Bridge
9.2.6. Setting Up 802.1q VLAN Tagging
9.2.7. Alias and Clone Files
9.2.8. Dialup Interfaces
9.2.9. Other Interfaces
9.3. Interface Control Scripts
9.4. Static Routes and the Default Gateway
9.5. Configuring IPv6 Tokenized Interface Identifiers
9.6. Network Function Files
9.7. Ethtool
9.8. Additional Resources
9.8.1. Installed Documentation
9.8.2. Useful Websites
10. Configure SCTP
10.1. Introduction to Streaming Control Transport Protocol (SCTP)
10.1.1. Comparison of TCP and SCTP Handshaking
10.2. Understanding SCTP
10.2.1. Bundled Streams
10.2.2. Partial Reliability
10.2.3. Message Boundary Preservation
10.2.4. Protocol Event Notifications
10.3. When To Use SCTP
10.4. When Not To Use SCTP
10.5. Compare SCTP to Bonding and Network Teaming
10.6. Using SCTP
10.6.1. Check if SCTP is installed
10.6.2. Install SCTP
10.6.3. Configuring SCTP
10.6.4. Tune SCTP
10.6.5. Troubleshooting SCTP
10.7. Additional Resources
10.7.1. Installed Documentation
10.7.2. Useful Websites

Chapter 8. NetworkManager

NetworkManager is a dynamic network control and configuration system that attempts to keep network devices and connections up and active when they are available. NetworkManager consists of a core daemon, a GNOME Notification Area applet that provides network status information, and graphical configuration tools that can create, edit and remove connections and interfaces. NetworkManager can be used to configure the following types of connections: Ethernet, wireless, mobile broadband (such as cellular 3G), and DSL and PPPoE (Point-to-Point over Ethernet). In addition, NetworkManager allows for the configuration of network aliases, static routes, DNS information and VPN connections, as well as many connection-specific parameters. Finally, NetworkManager provides a rich API via D-Bus which allows applications to query and control network configuration and state.
Previous versions of Red Hat Enterprise Linux included the Network Administration Tool, which was commonly known as system-config-network after its command line invocation. In Red Hat Enterprise Linux 6, NetworkManager replaces the former Network Administration Tool while providing enhanced functionality, such as user-specific and mobile broadband configuration. It is also possible to configure the network in Red Hat Enterprise Linux 6 by editing interface configuration files; refer to Chapter 9, Network Interfaces for more information.
NetworkManager may be installed by default on Red Hat Enterprise Linux. To ensure that it is, first run the following command as the root user:
~]# yum install NetworkManager

8.1. The NetworkManager Daemon

The NetworkManager daemon runs with root privileges and is usually configured to start up at boot time. You can determine whether the NetworkManager daemon is running by entering this command as root:
~]# service NetworkManager status
    NetworkManager (pid  1527) is running...
The service command will report NetworkManager is stopped if the NetworkManager service is not running. To start it for the current session:
~]# service NetworkManager start
Run the chkconfig command to ensure that NetworkManager starts up every time the system boots:
~]# chkconfig NetworkManager on
For more information on starting, stopping and managing services and runlevels, refer to Chapter 11, Services and Daemons.

8.2. Interacting with NetworkManager

Users do not interact with the NetworkManager system service directly. Instead, you can perform network configuration tasks via NetworkManager's Notification Area applet. The applet has multiple states that serve as visual indicators for the type of connection you are currently using. Hover the pointer over the applet icon for tooltip information on the current connection state.
NetworkManager applet states

Figure 8.1. NetworkManager applet states


If you do not see the NetworkManager applet in the GNOME panel, and assuming that the NetworkManager package is installed on your system, you can start the applet by running the following command as a normal user (not root):
~]$ nm-applet &
After running this command, the applet appears in your Notification Area. You can ensure that the applet runs each time you log in by clicking SystemPreferencesStartup Applications to open the Startup Applications Preferences window. Then, select the Startup Programs tab and check the box next to NetworkManager.

8.2.1. Connecting to a Network

When you left-click on the applet icon, you are presented with:
  • a list of categorized networks you are currently connected to (such as Wired and Wireless);
  • a list of all Available Networks that NetworkManager has detected;
  • options for connecting to any configured Virtual Private Networks (VPNs); and,
  • options for connecting to hidden or new wireless networks.
If you are connected to a network, its name is presented in bold typeface under its network type, such as Wired or Wireless. When many networks are available, such as wireless access points, the More networks expandable menu entry appears.
The NetworkManager applet's left-click menu, showing all available and connected-to networks

Figure 8.2. The NetworkManager applet's left-click menu, showing all available and connected-to networks


8.2.2. Configuring New and Editing Existing Connections

Next, right-click on the NetworkManager applet to open its context menu, which is the main point of entry for interacting with NetworkManager to configure connections.
The NetworkManager applet's context menu

Figure 8.3. The NetworkManager applet's context menu


Ensure that the Enable Networking box is checked. If the system has detected a wireless card, then you will also see an Enable Wireless menu option. Check the Enable Wireless checkbox as well. NetworkManager notifies you of network connection status changes if you check the Enable Notifications box. Clicking the Connection Information entry presents an informative Connection Information window that lists the connection type and interface, your IP address and routing details, and so on.
Finally, clicking on Edit Connections opens the Network Connections window, from where you can perform most of your network configuration tasks. Note that this window can also be opened by running, as a normal user:
~]$ nm-connection-editor &
Configure networks using the Network Connections window

Figure 8.4. Configure networks using the Network Connections window


There is an arrow head symbol to the left which can be clicked to hide and reveal entries as needed. To create a new connection, click the Add button to view the selection list, select the connection type and click the Create button. Alternatively, to edit an existing connection select the interface name from the list and click the Edit button.
Then, to configure:

8.2.3. Connecting to a Network Automatically

For any connection type you add or configure, you can choose whether you want NetworkManager to try to connect to that network automatically when it is available.

Procedure 8.1. Configuring NetworkManager to Connect to a Network Automatically When Detected

  1. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  2. Click the arrow head if necessary to reveal the list of connections.
  3. Select the specific connection that you want to configure and click Edit.
  4. Check Connect automatically to cause NetworkManager to auto-connect to the connection whenever NetworkManager detects that it is available. Uncheck the checkbox if you do not want NetworkManager to connect automatically. If the box is unchecked, you will have to select that connection manually in the NetworkManager applet's left-click menu to cause it to connect.

8.2.4. User and System Connections

NetworkManager connections are always either user connections or system connections. Depending on the system-specific policy that the administrator has configured, users may need root privileges to create and modify system connections. NetworkManager's default policy enables users to create and modify user connections, but requires them to have root privileges to add, modify or delete system connections.
User connections are so-called because they are specific to the user who creates them. In contrast to system connections, whose configurations are stored under the /etc/sysconfig/network-scripts/ directory (mainly in ifcfg-<network_type> interface configuration files), user connection settings are stored in the GConf configuration database and the GNOME keyring, and are only available during login sessions for the user who created them. Thus, logging out of the desktop session causes user-specific connections to become unavailable.

Increase security by making VPN connections user-specific

Because NetworkManager uses the GConf and GNOME keyring applications to store user connection settings, and because these settings are specific to your desktop session, it is highly recommended to configure your personal VPN connections as user connections. If you do so, other Non-root users on the system cannot view or access these connections in any way.
System connections, on the other hand, become available at boot time and can be used by other users on the system without first logging in to a desktop session.
NetworkManager can quickly and conveniently convert user to system connections and vice versa. Converting a user connection to a system connection causes NetworkManager to create the relevant interface configuration files under the /etc/sysconfig/network-scripts/ directory, and to delete the GConf settings from the user's session. Conversely, converting a system to a user-specific connection causes NetworkManager to remove the system-wide configuration files and create the corresponding GConf/GNOME keyring settings.
The Available to all users checkbox controls whether connections are user-specific or system-wide

Figure 8.5. The Available to all users checkbox controls whether connections are user-specific or system-wide


Procedure 8.2. Changing a Connection to be User-Specific instead of System-Wide, or Vice-Versa

Root privileges may be required

Depending on the system's policy, you may need root privileges on the system in order to change whether a connection is user-specific or system-wide.
  1. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  2. If needed, select the arrow head (on the left hand side) to hide and reveal the types of available network connections.
  3. Select the specific connection that you want to configure and click Edit.
  4. Check the Available to all users checkbox to ask NetworkManager to make the connection a system-wide connection. Depending on system policy, you may then be prompted for the root password by the PolicyKit application. If so, enter the root password to finalize the change.
    Conversely, uncheck the Available to all users checkbox to make the connection user-specific.

8.3. Establishing Connections

8.3.1. Establishing a Wired (Ethernet) Connection

To establish a wired network connection, Right-click on the NetworkManager applet to open its context menu, ensure that the Enable Networking box is checked, then click on Edit Connections. This opens the Network Connections window. Note that this window can also be opened by running, as a normal user:
~]$ nm-connection-editor &
You can click on the arrow head to reveal and hide the list of connections as needed.
The Network Connections window showing the newly created System eth0 connection

Figure 8.6. The Network Connections window showing the newly created System eth0 connection


The system startup scripts create and configure a single wired connection called System eth0 by default on all systems. Although you can edit System eth0, creating a new wired connection for your custom settings is recommended. You can create a new wired connection by clicking the Add button, selecting the Wired entry from the list that appears and then clicking the Create button.
Selecting a new connection type from the "Choose a Connection Type" list

Figure 8.7. Selecting a new connection type from the "Choose a Connection Type" list


The dialog for adding and editing connections is the same

When you add a new connection by clicking the Add button, a list of connection types appears. Once you have made a selection and clicked on the Create button, NetworkManager creates a new configuration file for that connection and then opens the same dialog that is used for editing an existing connection. There is no difference between these dialogs. In effect, you are always editing a connection; the difference only lies in whether that connection previously existed or was just created by NetworkManager when you clicked Create.
Editing the newly created Wired connection System eth0

Figure 8.8.  Editing the newly created Wired connection System eth0


Configuring the Connection Name, Auto-Connect Behavior, and Availability Settings

Three settings in the Editing dialog are common to all connection types:
  • Connection name — Enter a descriptive name for your network connection. This name will be used to list this connection in the Wired section of the Network Connections window.
  • Connect automatically — Check this box if you want NetworkManager to auto-connect to this connection when it is available. Refer to Section 8.2.3, “Connecting to a Network Automatically” for more information.
  • Available to all users — Check this box to create a connection available to all users on the system. Changing this setting may require root privileges. Refer to Section 8.2.4, “User and System Connections” for details.

Configuring the Wired Tab

The final three configurable settings are located within the Wired tab itself: the first is a text-entry field where you can specify a MAC (Media Access Control) address, and the second allows you to specify a cloned MAC address, and third allows you to specify the MTU (Maximum Transmission Unit) value. Normally, you can leave the MAC address field blank and the MTU set to automatic. These defaults will suffice unless you are associating a wired connection with a second or specific NIC, or performing advanced networking. In such cases, refer to the following descriptions:
MAC Address
Network hardware such as a Network Interface Card (NIC) has a unique MAC address (Media Access Control; also known as a hardware address) that identifies it to the system. Running the ip addr command will show the MAC address associated with each interface. For example, in the following ip addr output, the MAC address for the eth0 interface (which is 52:54:00:26:9e:f1) immediately follows the link/ether keyword:
~]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UNKNOWN qlen 1000
    link/ether 52:54:00:26:9e:f1 brd ff:ff:ff:ff:ff:ff
    inet 192.168.122.251/24 brd 192.168.122.255 scope global eth0
    inet6 fe80::5054:ff:fe26:9ef1/64 scope link
       valid_lft forever preferred_lft forever
A single system can have one or more NICs installed on it. The MAC address field therefore allows you to associate a specific NIC with a specific connection (or connections). As mentioned, you can determine the MAC address using the ip addr command, and then copy and paste that value into the MAC address text-entry field.
The cloned MAC address field is mostly for use in such situations were a network service has been restricted to a specific MAC address and you need to emulate that MAC address.
MTU
The MTU (Maximum Transmission Unit) value represents the size in bytes of the largest packet that the connection will use to transmit. This value defaults to 1500 when using IPv4, or a variable number 1280 or higher for IPv6, and does not generally need to be specified or changed.

Saving Your New (or Modified) Connection and Making Further Configurations

Once you have finished editing your wired connection, click the Apply button and NetworkManager will immediately save your customized configuration. Given a correct configuration, you can connect to your new or customized connection by selecting it from the NetworkManager Notification Area applet. See Section 8.2.1, “Connecting to a Network” for information on using your new or altered connection.
You can further configure an existing connection by selecting it in the Network Connections window and clicking Edit to return to the Editing dialog.
Then, to configure:

8.3.2. Establishing a Wireless Connection

This section explains how to use NetworkManager to configure a wireless (also known as Wi-Fi or 802.1a/b/g/n) connection to an Access Point.
To configure a mobile broadband (such as 3G) connection, refer to Section 8.3.3, “Establishing a Mobile Broadband Connection”.

Quickly Connecting to an Available Access Point

The easiest way to connect to an available access point is to left-click on the NetworkManager applet, locate the Service Set Identifier (SSID) of the access point in the list of Available networks, and click on it. If the access point is secured, a dialog prompts you for authentication.
Authenticating to a wireless access point

Figure 8.9. Authenticating to a wireless access point


NetworkManager tries to auto-detect the type of security used by the access point. If there are multiple possibilities, NetworkManager guesses the security type and presents it in the Wireless security dropdown menu. To see if there are multiple choices, click the Wireless security dropdown menu and select the type of security the access point is using. If you are unsure, try connecting to each type in turn. Finally, enter the key or passphrase in the Password field. Certain password types, such as a 40-bit WEP or 128-bit WPA key, are invalid unless they are of a requisite length. The Connect button will remain inactive until you enter a key of the length required for the selected security type. To learn more about wireless security, refer to Section 8.3.9.2, “Configuring Wireless Security”.

Prevent Roaming On The Same Access Point

In the case of WPA and WPA2 (Personal and Enterprise), an option to select between Auto, WPA and WPA2 has been added. This option is intended for use with an access point that is offering both WPA and WPA2. Select one of the protocols if you would like to prevent roaming between the two protocols. Roaming between WPA and WPA2 on the same access point can cause loss of service.
If NetworkManager connects to the access point successfully, its applet icon will change into a graphical indicator of the wireless connection's signal strength.
Applet icon indicating a wireless connection signal strength of 75%

Figure 8.10. Applet icon indicating a wireless connection signal strength of 75%


You can also edit the settings for one of these auto-created access point connections just as if you had added it yourself. The Wireless tab of the Network Connections window lists all of the connections you have ever tried to connect to: NetworkManager names each of them Auto <SSID>, where SSID is the Service Set identifier of the access point.
An example of access points that have previously been connected to

Figure 8.11. An example of access points that have previously been connected to


Connecting to a Hidden Wireless Network

All access points have a Service Set Identifier (SSID) to identify them. However, an access point may be configured not to broadcast its SSID, in which case it is hidden, and will not show up in NetworkManager's list of Available networks. You can still connect to a wireless access point that is hiding its SSID as long as you know its SSID, authentication method, and secrets.
To connect to a hidden wireless network, left-click NetworkManager's applet icon and select Connect to Hidden Wireless Network... to cause a dialog to appear. If you have connected to the hidden network before, use the Connection dropdown to select it, and click Connect. If you have not, leave the Connection dropdown as New..., enter the SSID of the hidden network, select its Wireless security method, enter the correct authentication secrets, and click Connect.
For more information on wireless security settings, refer to Section 8.3.9.2, “Configuring Wireless Security”.

Editing a Connection, or Creating a Completely New One

You can edit an existing connection that you have tried or succeeded in connecting to in the past by opening the Wireless tab of the Network Connections, selecting the connection by name (words which follow Auto refer to the SSID of an access point), and clicking Edit.
You can create a new connection by opening the Network Connections window, clicking the Add button, selecting Wireless, and clicking the Create button.
  1. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  2. Click the Add button.
  3. Select the Wireless entry from the list.
  4. Click the Create button.
Editing the newly created Wireless connection 1

Figure 8.12. Editing the newly created Wireless connection 1


Configuring the Connection Name, Auto-Connect Behavior, and Availability Settings

Three settings in the Editing dialog are common to all connection types:
  • Connection name — Enter a descriptive name for your network connection. This name will be used to list this connection in the Wireless section of the Network Connections window. By default, wireless connections are named the same as the SSID of the wireless access point. You can rename the wireless connection without affecting its ability to connect, but it is recommended to retain the SSID name.
  • Connect automatically — Check this box if you want NetworkManager to auto-connect to this connection when it is available. Refer to Section 8.2.3, “Connecting to a Network Automatically” for more information.
  • Available to all users — Check this box to create a connection available to all users on the system. Changing this setting may require root privileges. Refer to Section 8.2.4, “User and System Connections” for details.

Configuring the Wireless Tab

SSID
All access points have a Service Set identifier to identify them. However, an access point may be configured not to broadcast its SSID, in which case it is hidden, and will not show up in NetworkManager's list of Available networks. You can still connect to a wireless access point that is hiding its SSID as long as you know its SSID (and authentication secrets).
For information on connecting to a hidden wireless network, refer to Section 8.3.2, “Connecting to a Hidden Wireless Network”.
Mode
Infrastructure — Set Mode to Infrastructure if you are connecting to a dedicated wireless access point or one built into a network device such as a router or a switch.
Ad-hoc — Set Mode to Ad-hoc if you are creating a peer-to-peer network for two or more mobile devices to communicate directly with each other. If you use Ad-hoc mode, referred to as Independent Basic Service Set (IBSS) in the 802.11 standard, you must ensure that the same SSID is set for all participating wireless devices, and that they are all communicating over the same channel.
BSSID
The Basic Service Set Identifier (BSSID) is the MAC address of the specific wireless access point you are connecting to when in Infrastructure mode. This field is blank by default, and you are able to connect to a wireless access point by SSID without having to specify its BSSID. If the BSSID is specified, it will force the system to associate to a specific access point only.
For ad-hoc networks, the BSSID is generated randomly by the mac80211 subsystem when the ad-hoc network is created. It is not displayed by NetworkManager
MAC address
Like an Ethernet Network Interface Card (NIC), a wireless adapter has a unique MAC address (Media Access Control; also known as a hardware address) that identifies it to the system. Running the ip addr command will show the MAC address associated with each interface. For example, in the following ip addr output, the MAC address for the wlan0 interface (which is 00:1c:bf:02:f8:70) immediately follows the link/ether keyword:
~]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UNKNOWN qlen 1000
    link/ether 52:54:00:26:9e:f1 brd ff:ff:ff:ff:ff:ff
    inet 192.168.122.251/24 brd 192.168.122.255 scope global eth0
    inet6 fe80::5054:ff:fe26:9ef1/64 scope link
       valid_lft forever preferred_lft forever
3: wlan0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP qlen 1000
    link/ether 00:1c:bf:02:f8:70 brd ff:ff:ff:ff:ff:ff
    inet 10.200.130.67/24 brd 10.200.130.255 scope global wlan0
    inet6 fe80::21c:bfff:fe02:f870/64 scope link
       valid_lft forever preferred_lft forever
A single system could have one or more wireless network adapters connected to it. The MAC address field therefore allows you to associate a specific wireless adapter with a specific connection (or connections). As mentioned, you can determine the MAC address using the ip addr command, and then copy and paste that value into the MAC address text-entry field.
MTU
The MTU (Maximum Transmission Unit) value represents the size in bytes of the largest packet that the connection will use to transmit. If set to a non-zero number, only packets of the specified size or smaller will be transmitted. Larger packets are broken up into multiple Ethernet frames. It is recommended to leave this setting on automatic.

Saving Your New (or Modified) Connection and Making Further Configurations

Once you have finished editing the wireless connection, click the Apply button and NetworkManager will immediately save your customized configuration. Given a correct configuration, you can successfully connect to your the modified connection by selecting it from the NetworkManager Notification Area applet. See Section 8.2.1, “Connecting to a Network” for details on selecting and connecting to a network.
You can further configure an existing connection by selecting it in the Network Connections window and clicking Edit to return to the Editing dialog.
Then, to configure:

8.3.3. Establishing a Mobile Broadband Connection

You can use NetworkManager's mobile broadband connection abilities to connect to the following 2G and 3G services:
  • 2G — GPRS (General Packet Radio Service) or EDGE (Enhanced Data Rates for GSM Evolution)
  • 3G — UMTS (Universal Mobile Telecommunications System) or HSPA (High Speed Packet Access)
Your computer must have a mobile broadband device (modem), which the system has discovered and recognized, in order to create the connection. Such a device may be built into your computer (as is the case on many notebooks and netbooks), or may be provided separately as internal or external hardware. Examples include PC card, USB Modem or Dongle, mobile or cellular telephone capable of acting as a modem.

Procedure 8.3. Adding a New Mobile Broadband Connection

You can configure a mobile broadband connection by opening the Network Connections window, clicking Add, and selecting Mobile Broadband from the list.
  1. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  2. Click the Add button to open the selection list. Select Mobile Broadband and then click Create. The Set up a Mobile Broadband Connection assistant appears.
  3. Under Create a connection for this mobile broadband device, choose the 2G- or 3G-capable device you want to use with the connection. If the dropdown menu is inactive, this indicates that the system was unable to detect a device capable of mobile broadband. In this case, click Cancel, ensure that you do have a mobile broadband-capable device attached and recognized by the computer and then retry this procedure. Click the Forward button.
  4. Select the country where your service provider is located from the list and click the Forward button.
  5. Select your provider from the list or enter it manually. Click the Forward button.
  6. Select your payment plan from the dropdown menu and confirm the Access Point Name (APN) is correct. Click the Forward button.
  7. Review and confirm the settings and then click the Apply button.
  8. Edit the mobile broadband-specific settings by referring to the Configuring the Mobile Broadband Tab description below .

Procedure 8.4. Editing an Existing Mobile Broadband Connection

Follow these steps to edit an existing mobile broadband connection.
  1. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  2. Select the connection you wish to edit and click the Edit button.
  3. Select the Mobile Broadband tab.
  4. Configure the connection name, auto-connect behavior, and availability settings.
    Three settings in the Editing dialog are common to all connection types:
    • Connection name — Enter a descriptive name for your network connection. This name will be used to list this connection in the Mobile Broadband section of the Network Connections window.
    • Connect automatically — Check this box if you want NetworkManager to auto-connect to this connection when it is available. Refer to Section 8.2.3, “Connecting to a Network Automatically” for more information.
    • Available to all users — Check this box to create a connection available to all users on the system. Changing this setting may require root privileges. Refer to Section 8.2.4, “User and System Connections” for details.
  5. Edit the mobile broadband-specific settings by referring to the Configuring the Mobile Broadband Tab description below .

Saving Your New (or Modified) Connection and Making Further Configurations

Once you have finished editing your mobile broadband connection, click the Apply button and NetworkManager will immediately save your customized configuration. Given a correct configuration, you can connect to your new or customized connection by selecting it from the NetworkManager Notification Area applet. See Section 8.2.1, “Connecting to a Network” for information on using your new or altered connection.
You can further configure an existing connection by selecting it in the Network Connections window and clicking Edit to return to the Editing dialog.
Then, to configure:

Configuring the Mobile Broadband Tab

If you have already added a new mobile broadband connection using the assistant (refer to Procedure 8.3, “Adding a New Mobile Broadband Connection” for instructions), you can edit the Mobile Broadband tab to disable roaming if home network is not available, assign a network ID, or instruct NetworkManager to prefer a certain technology (such as 3G or 2G) when using the connection.
Number
The number that is dialed to establish a PPP connection with the GSM-based mobile broadband network. This field may be automatically populated during the initial installation of the broadband device. You can usually leave this field blank and enter the APN instead.
Username
Enter the user name used to authenticate with the network. Some providers do not provide a user name, or accept any user name when connecting to the network.
Password
Enter the password used to authenticate with the network. Some providers do not provide a password, or accept any password.
APN
Enter the Access Point Name (APN) used to establish a connection with the GSM-based network. Entering the correct APN for a connection is important because it often determines:
  • how the user is billed for their network usage; and/or
  • whether the user has access to the Internet, an intranet, or a subnetwork.
Network ID
Entering a Network ID causes NetworkManager to force the device to register only to a specific network. This can be used to ensure the connection does not roam when it is not possible to control roaming directly.
Type
Any — The default value of Any leaves the modem to select the fastest network.
3G (UMTS/HSPA) — Force the connection to use only 3G network technologies.
2G (GPRS/EDGE) — Force the connection to use only 2G network technologies.
Prefer 3G (UMTS/HSPA) — First attempt to connect using a 3G technology such as HSPA or UMTS, and fall back to GPRS or EDGE only upon failure.
Prefer 2G (GPRS/EDGE) — First attempt to connect using a 2G technology such as GPRS or EDGE, and fall back to HSPA or UMTS only upon failure.
Allow roaming if home network is not available
Uncheck this box if you want NetworkManager to terminate the connection rather than transition from the home network to a roaming one, thereby avoiding possible roaming charges. If the box is checked, NetworkManager will attempt to maintain a good connection by transitioning from the home network to a roaming one, and vice versa.
PIN
If your device's SIM (Subscriber Identity Module) is locked with a PIN (Personal Identification Number), enter the PIN so that NetworkManager can unlock the device. NetworkManager must unlock the SIM if a PIN is required in order to use the device for any purpose.

8.3.4. Establishing a VPN Connection

Establishing an encrypted Virtual Private Network (VPN) enables you to communicate securely between your Local Area Network (LAN), and another, remote LAN. After successfully establishing a VPN connection, a VPN router or gateway performs the following actions upon the packets you transmit:
  1. it adds an Authentication Header for routing and authentication purposes;
  2. it encrypts the packet data; and,
  3. it encloses the data with an Encapsulating Security Payload (ESP), which constitutes the decryption and handling instructions.
The receiving VPN router strips the header information, decrypts the data, and routes it to its intended destination (either a workstation or other node on a network). Using a network-to-network connection, the receiving node on the local network receives the packets already decrypted and ready for processing. The encryption/decryption process in a network-to-network VPN connection is therefore transparent to clients.
Because they employ several layers of authentication and encryption, VPNs are a secure and effective means of connecting multiple remote nodes to act as a unified intranet.

Procedure 8.5. Adding a New VPN Connection

  1. You can configure a new VPN connection by opening the Network Connections window, clicking the Add button and selecting a type of VPN from the VPN section of the new connection list.
  2. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  3. Click the Add button.
  4. The Choose a Connection Type list appears.
  5. A VPN plug-in is required

    The appropriate NetworkManager VPN plug-in for the VPN type you want to configure must be installed. (refer to Section 6.2.4, “Installing Packages” for more information on how to install new packages in Red Hat Enterprise Linux 6). The VPN section in the Choose a Connection Type list will not appear if you do not have a suitable plug-in installed.
  6. Select the VPN protocol for the gateway you are connecting to from the Choose a Connection Type list. The VPN protocols available for selection in the list correspond to the NetworkManager VPN plug-ins installed. For example, if the NetworkManager VPN plug-in for openswan is installed then the IPsec based VPN will be selectable from the Choose a Connection Type list.
    After selecting the correct one, press the Create button.
  7. The Editing VPN Connection 1 window then appears. This window presents settings customized for the type of VPN connection you selected in Step 6.

Procedure 8.6. Editing an Existing VPN Connection

You can configure an existing VPN connection by opening the Network Connections window and selecting the name of the connection from the list. Then click the Edit button.
  1. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  2. Select the connection you wish to edit and click the Edit button.
Editing the newly created IPsec VPN connection 1

Figure 8.13. Editing the newly created IPsec VPN connection 1


Configuring the Connection Name, Auto-Connect Behavior, and Availability Settings

Three settings in the Editing dialog are common to all connection types:
  • Connection name — Enter a descriptive name for your network connection. This name will be used to list this connection in the VPN section of the Network Connections window.
  • Connect automatically — Check this box if you want NetworkManager to auto-connect to this connection when it is available. Refer to Section 8.2.3, “Connecting to a Network Automatically” for more information.
  • Available to all users — Check this box to create a connection available to all users on the system. Changing this setting may require root privileges. Refer to Section 8.2.4, “User and System Connections” for details.

Configuring the VPN Tab

Gateway
The name or IP address of the remote VPN gateway.
Group name
The name of a VPN group configured on the remote gateway.
User password
If required, enter the password used to authenticate with the VPN.
Group password
If required, enter the password used to authenticate with the VPN.
User name
If required, enter the user name used to authenticate with the VPN.
Phase1 Algorithms
If required, enter the algorithms to be used to authenticate and set up an encrypted channel.
Phase2 Algorithms
If required, enter the algorithms to be used for the IPsec negotiations.
Domain
If required, enter the Domain Name.
NAT traversal
Cisco UDP (default) — IPsec over UDP.
NAT-T — ESP encapsulation and IKE extensions are used to handle NAT Traversal.
Disabled — No special NAT measures required.
Disable Dead Peer Detection — Disable the sending of probes to the remote gateway or endpoint.

Saving Your New (or Modified) Connection and Making Further Configurations

Once you have finished editing your new VPN connection, click the Apply button and NetworkManager will immediately save your customized configuration. Given a correct configuration, you can connect to your new or customized connection by selecting it from the NetworkManager Notification Area applet. See Section 8.2.1, “Connecting to a Network” for information on using your new or altered connection.
You can further configure an existing connection by selecting it in the Network Connections window and clicking Edit to return to the Editing dialog.
Then, to configure:

8.3.5. Establishing a DSL Connection

This section is intended for those installations which have a DSL card fitted within a host rather than the external combined DSL modem router combinations typical of private consumer or SOHO installations.

Procedure 8.7. Adding a New DSL Connection

You can configure a new DSL connection by opening the Network Connections window, clicking the Add button and selecting DSL from the Hardware section of the new connection list.
  1. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  2. Click the Add button.
  3. The Choose a Connection Type list appears.
  4. Select DSL and press the Create button.
  5. The Editing DSL Connection 1 window appears.

Procedure 8.8. Editing an Existing DSL Connection

You can configure an existing DSL connection by opening the Network Connections window and selecting the name of the connection from the list. Then click the Edit button.
  1. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  2. Select the connection you wish to edit and click the Edit button.

Configuring the Connection Name, Auto-Connect Behavior, and Availability Settings

Three settings in the Editing dialog are common to all connection types:
  • Connection name — Enter a descriptive name for your network connection. This name will be used to list this connection in the DSL section of the Network Connections window.
  • Connect automatically — Check this box if you want NetworkManager to auto-connect to this connection when it is available. Refer to Section 8.2.3, “Connecting to a Network Automatically” for more information.
  • Available to all users — Check this box to create a connection available to all users on the system. Changing this setting may require root privileges. Refer to Section 8.2.4, “User and System Connections” for details.

Configuring the DSL Tab

Username
Enter the user name used to authenticate with the service provider.
Service
Leave blank unless otherwise directed.
Password
Enter the password supplied by the service provider.

Saving Your New (or Modified) Connection and Making Further Configurations

Once you have finished editing your DSL connection, click the Apply button and NetworkManager will immediately save your customized configuration. Given a correct configuration, you can connect to your new or customized connection by selecting it from the NetworkManager Notification Area applet. See Section 8.2.1, “Connecting to a Network” for information on using your new or altered connection.
You can further configure an existing connection by selecting it in the Network Connections window and clicking Edit to return to the Editing dialog.
Then, to configure:

8.3.6. Establishing a Bond Connection

You can use NetworkManager to create a Bond from two or more Wired or Infiniband connections. It is not necessary to create the connections to be bonded first. They can be configured as part of the process to configure the bond. You must have the MAC addresses of the interfaces available in order to complete the configuration process.

Note

NetworkManager support for bonding must be enabled by means of the NM_BOND_VLAN_ENABLED directive and then NetworkManager must be restarted. See Section 9.2.1, “Ethernet Interfaces” for an explanation of NM_CONTROLLED and the NM_BOND_VLAN_ENABLED directive. See Section 11.3.4, “Restarting a Service” for an explantion of restarting a service such as NetworkManager from the command line. Alternatively, for a graphical tool see Section 11.2.1, “Using the Service Configuration Utility”.

Procedure 8.9. Adding a New Bond Connection

You can configure a Bond connection by opening the Network Connections window, clicking Add, and selecting Bond from the list.
  1. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  2. Click the Add button to open the selection list. Select Bond and then click Create. The Editing Bond Connection 1 window appears.
  3. On the Bond tab, click Add and select the type of interface you want to use with the bond connection. Click the Create button. Note that the dialog to select the slave type only comes up when you create the first slave; after that, it will automatically use that same type for all further slaves.
  4. The Editing bond1 slave1 window appears. Fill in the MAC address of the first interface to be bonded. Click the Apply button.
  5. The Authenticate window appears. Enter the root password to continue. Click the Authenticate button.
  6. The name of the bonded slave appears in the Bonded Connections window. Click the Add button to add further slave connections.
  7. Review and confirm the settings and then click the Apply button.
  8. Edit the bond-specific settings by referring to Section 8.3.6, “Configuring the Bond Tab” below.
Editing the newly created Bond connection 1

Figure 8.14. Editing the newly created Bond connection 1


Procedure 8.10. Editing an Existing Bond Connection

Follow these steps to edit an existing bond connection.
  1. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  2. Select the connection you wish to edit and click the Edit button.
  3. Select the Bond tab.
  4. Configure the connection name, auto-connect behavior, and availability settings.
    Three settings in the Editing dialog are common to all connection types:
    • Connection name — Enter a descriptive name for your network connection. This name will be used to list this connection in the Bond section of the Network Connections window.
    • Connect automatically — Check this box if you want NetworkManager to auto-connect to this connection when it is available. Refer to Section 8.2.3, “Connecting to a Network Automatically” for more information.
    • Available to all users — Check this box to create a connection available to all users on the system. Changing this setting may require root privileges. Refer to Section 8.2.4, “User and System Connections” for details.
  5. Edit the bond-specific settings by referring to Section 8.3.6, “Configuring the Bond Tab” below.

Saving Your New (or Modified) Connection and Making Further Configurations

Once you have finished editing your bond connection, click the Apply button and NetworkManager will immediately save your customized configuration. Given a correct configuration, you can connect to your new or customized connection by selecting it from the NetworkManager Notification Area applet. See Section 8.2.1, “Connecting to a Network” for information on using your new or altered connection.
You can further configure an existing connection by selecting it in the Network Connections window and clicking Edit to return to the Editing dialog.
Then, to configure:

Configuring the Bond Tab

If you have already added a new bond connection (refer to Procedure 8.9, “Adding a New Bond Connection” for instructions), you can edit the Bond tab to set the load sharing mode and the type of link monitoring to use to detect failures of a slave connection.
Mode
The mode that is used to share traffic over the slave connections which make up the bond. The default is Round-robin. Other load sharing modes, such as 802.3ad, may be selected by means of the drop-down list.
Link Monitoring
The method of monitoring the slaves ability to carry network traffic.
The following modes of load sharing are selectable from the Mode drop-down list:
Round-robin
Sets a round-robin policy for fault tolerance and load balancing. Transmissions are received and sent out sequentially on each bonded slave interface beginning with the first one available.
Active backup
Sets an active-backup policy for fault tolerance. Transmissions are received and sent out via the first available bonded slave interface. Another bonded slave interface is only used if the active bonded slave interface fails. Note that this is the only mode available for bonds of InfiniBand devices.
XOR
Sets an XOR (exclusive-or) policy for fault tolerance and load balancing. Using this method, the interface matches up the incoming request's MAC address with the MAC address for one of the slave NICs. Once this link is established, transmissions are sent out sequentially beginning with the first available interface.
Broadcast
Sets a broadcast policy for fault tolerance. All transmissions are sent on all slave interfaces.
802.3ad
Sets an IEEE 802.3ad dynamic link aggregation policy. Creates aggregation groups that share the same speed and duplex settings. Transmits and receives on all slaves in the active aggregator. Requires a switch that is 802.3ad compliant.
Adaptive transmit load balancing
Sets an adaptive Transmit Load Balancing (TLB) policy for fault tolerance and load balancing. The outgoing traffic is distributed according to the current load on each slave interface. Incoming traffic is received by the current slave. If the receiving slave fails, another slave takes over the MAC address of the failed slave.
Active Load Balancing
Sets an Active Load Balancing (ALB) policy for fault tolerance and load balancing. Includes transmit and receive load balancing for IPv4 traffic. Receive load balancing is achieved through ARP negotiation.
The following types of link monitoring can be selected from the Link Monitoring drop-down list. It is a good idea to test which channel bonding module parameters work best for your bonded interfaces.
MII (Media Independent Interface)
The state of the carrier wave of the interface is monitored. This can be done by querying the driver, by querying MII registers directly, or by using Ethtool to query the device. Three options are available:
Monitoring Frequency
The time interval, in milliseconds, between querying the driver or MII registers.
Link up delay
The time in milleseconds to wait before attempting to use a link that has been reported as up. This delay can be used if some gratuitous ARP requests are lost in the period immediately following the link being reported as up. This can happen during switch initialization for example.
Link down delay
The time in milleseconds to wait before changing to another link when a previously active link has been reported as down. This delay can be used if an attached switch takes a relatively long time to change to backup mode.
ARP
The Address Resolution Protocol (ARP) is used to probe one or more peers to determine how well the link layer connections are working. It is dependent on the device driver providing the transmit start time and the last receive time. Two options are available:
Monitoring Frequency
The time interval, in milliseconds, between sending ARP requests.
ARP targets
A comma separated list of IP addresses to send ARP requests to.

8.3.7. Establishing a VLAN Connection

You can use NetworkManager to create a VLAN using an existing interface. Currently, at time of writing, you can only make VLANs on Ethernet devices.

Procedure 8.11. Adding a New VLAN Connection

You can configure a VLAN connection by opening the Network Connections window, clicking Add, and selecting VLAN from the list.
  1. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  2. Click the Add button to open the selection list. Select VLAN and then click Create. The Editing VLAN Connection 1 window appears.
  3. On the VLAN tab, select the parent interface from the drop-down list you want to use for the VLAN connection.
  4. Enter the VLAN ID
  5. Enter a VLAN interface name. This is the name of the VLAN interface that will be created. For example, "eth0.1" or "vlan2". (Normally this is either the parent interface name plus "." and the VLAN ID, or "vlan" plus the VLAN ID.)
  6. Review and confirm the settings and then click the Apply button.
  7. Edit the VLAN-specific settings by referring to the Configuring the VLAN Tab description below .

Procedure 8.12. Editing an Existing VLAN Connection

Follow these steps to edit an existing VLAN connection.
  1. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  2. Select the connection you wish to edit and click the Edit button.
  3. Select the VLAN tab.
  4. Configure the connection name, auto-connect behavior, and availability settings.
    Three settings in the Editing dialog are common to all connection types:
    • Connection name — Enter a descriptive name for your network connection. This name will be used to list this connection in the VLAN section of the Network Connections window.
    • Connect automatically — Check this box if you want NetworkManager to auto-connect to this connection when it is available. Refer to Section 8.2.3, “Connecting to a Network Automatically” for more information.
    • Available to all users — Check this box to create a connection available to all users on the system. Changing this setting may require root privileges. Refer to Section 8.2.4, “User and System Connections” for details.
  5. Edit the VLAN-specific settings by referring to the Configuring the VLAN Tab description below .

Saving Your New (or Modified) Connection and Making Further Configurations

Once you have finished editing your VLAN connection, click the Apply button and NetworkManager will immediately save your customized configuration. Given a correct configuration, you can connect to your new or customized connection by selecting it from the NetworkManager Notification Area applet. See Section 8.2.1, “Connecting to a Network” for information on using your new or altered connection.
You can further configure an existing connection by selecting it in the Network Connections window and clicking Edit to return to the Editing dialog.
Then, to configure:

Configuring the VLAN Tab

If you have already added a new VLAN connection (refer to Procedure 8.11, “Adding a New VLAN Connection” for instructions), you can edit the VLAN tab to set the parent interface and the VLAN ID.
Parent Interface
A previously configured interface can be selected in the drop-down list.
VLAN ID
The identification number to be used to tag the VLAN network traffic.
VLAN interface name
The name of the VLAN interface that will be created. For example, "eth0.1" or "vlan2".
Cloned MAC address
Optionally sets an alternate MAC address to use for identifying the VLAN interface. This can be used to change the source MAC address for packets sent on this VLAN.
MTU
Optionally sets a Maximum Transmission Unit (MTU) size to be used for packets to be sent over the VLAN connection.

8.3.8. Establishing an IP-over-InfiniBand (IPoIB) Connection

You can use NetworkManager to create an InfiniBand connection.

Procedure 8.13. Adding a New InfiniBand Connection

You can configure an InfiniBand connection by opening the Network Connections window, clicking Add, and selecting InfiniBand from the list.
  1. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  2. Click the Add button to open the selection list. Select InfiniBand and then click Create. The Editing InfiniBand Connection 1 window appears.
  3. On the InfiniBand tab, select the transport mode from the drop-down list you want to use for the InfiniBand connection.
  4. Enter the InfiniBand MAC address.
  5. Review and confirm the settings and then click the Apply button.
  6. Edit the InfiniBand-specific settings by referring to the Configuring the InfiniBand Tab description below .
Editing the newly created InfiniBand connection 1

Figure 8.15. Editing the newly created InfiniBand connection 1


Procedure 8.14. Editing an Existing InfiniBand Connection

Follow these steps to edit an existing InfiniBand connection.
  1. Right-click on the NetworkManager applet icon in the Notification Area and click Edit Connections. The Network Connections window appears.
  2. Select the connection you wish to edit and click the Edit button.
  3. Select the InfiniBand tab.
  4. Configure the connection name, auto-connect behavior, and availability settings.
    Three settings in the Editing dialog are common to all connection types:
    • Connection name — Enter a descriptive name for your network connection. This name will be used to list this connection in the InfiniBand section of the Network Connections window.
    • Connect automatically — Check this box if you want NetworkManager to auto-connect to this connection when it is available. Refer to Section 8.2.3, “Connecting to a Network Automatically” for more information.
    • Available to all users — Check this box to create a connection available to all users on the system. Changing this setting may require root privileges. Refer to Section 8.2.4, “User and System Connections” for details.
  5. Edit the InfiniBand-specific settings by referring to the Configuring the InfiniBand Tab description below .

Saving Your New (or Modified) Connection and Making Further Configurations

Once you have finished editing your InfiniBand connection, click the Apply button and NetworkManager will immediately save your customized configuration. Given a correct configuration, you can connect to your new or customized connection by selecting it from the NetworkManager Notification Area applet. See Section 8.2.1, “Connecting to a Network” for information on using your new or altered connection.
You can further configure an existing connection by selecting it in the Network Connections window and clicking Edit to return to the Editing dialog.
Then, to configure:

Configuring the InfiniBand Tab

If you have already added a new InfiniBand connection (refer to Procedure 8.13, “Adding a New InfiniBand Connection” for instructions), you can edit the InfiniBand tab to set the parent interface and the InfiniBand ID.
Transport mode
Datagram or Connected mode can be selected from the drop-down list. Select the same mode the rest of your IPoIB network is using.
Device MAC address
The MAC address of the InfiniBand capable device to be used for the InfiniBand network traffic.This hardware address field will be pre-filled if you have InfiniBand hardware installed.
MTU
Optionally sets a Maximum Transmission Unit (MTU) size to be used for packets to be sent over the InfiniBand connection.

8.3.9. Configuring Connection Settings

8.3.9.1. Configuring 802.1x Security

802.1x security is the name of the IEEE standard for port-based Network Access Control (PNAC). Simply put, 802.1x security is a way of defining a logical network out of a physical one. All clients who want to join the logical network must authenticate with the server (a router, for example) using the correct 802.1x authentication method.
802.1x security is most often associated with securing wireless networks (WLANs), but can also be used to prevent intruders with physical access to the network (LAN) from gaining entry. In the past, DHCP servers were configured not to lease IP addresses to unauthorized users, but for various reasons this practice is both impractical and insecure, and thus is no longer recommended. Instead, 802.1x security is used to ensure a logically-secure network through port-based authentication.
802.1x provides a framework for WLAN and LAN access control and serves as an envelope for carrying one of the Extensible Authentication Protocol (EAP) types. An EAP type is a protocol that defines how WLAN security is achieved on the network.
You can configure 802.1x security for a wired or wireless connection type by opening the Network Connections window (refer to Section 8.2.2, “Configuring New and Editing Existing Connections”) and following the applicable procedure:

Procedure 8.15. For a wired connection...

  1. Either click Add, select a new network connection for which you want to configure 802.1x security and then click Create, or select an existing connection and click Edit.
  2. Then select the 802.1x Security tab and check the Use 802.1x security for this connection checkbox to enable settings configuration.

Procedure 8.16. For a wireless connection...

  1. Either click on Add, select a new network connection for which you want to configure 802.1x security and then click Create, or select an existing connection and click Edit.
  2. Select the Wireless Security tab.
  3. Then click the Security dropdown and choose one of the following security methods: LEAP, Dynamic WEP (802.1x), or WPA & WPA2 Enterprise.
  4. Refer to Section 8.3.9.1.1, “Configuring TLS (Transport Layer Security) Settings” for descriptions of which EAP types correspond to your selection in the Security dropdown.
8.3.9.1.1. Configuring TLS (Transport Layer Security) Settings
With Transport Layer Security, the client and server mutually authenticate using the TLS protocol. The server demonstrates that it holds a digital certificate, the client proves its own identity using its client-side certificate, and key information is exchanged. Once authentication is complete, the TLS tunnel is no longer used. Instead, the client and server use the exchanged keys to encrypt data using AES, TKIP or WEP.
The fact that certificates must be distributed to all clients who want to authenticate means that the EAP-TLS authentication method is very strong, but also more complicated to set up. Using TLS security requires the overhead of a public key infrastructure (PKI) to manage certificates. The benefit of using TLS security is that a compromised password does not allow access to the (W)LAN: an intruder must also have access to the authenticating client's private key.
NetworkManager does not determine the version of TLS supported. NetworkManager gathers the parameters entered by the user and passes them to the daemon, wpa_supplicant, that handles the procedure. It in turn uses OpenSSL to establish the TLS tunnel. OpenSSL itself negotiates the SSL/TLS protocol version. It uses the highest version both ends support.
Identity
Identity string for EAP authentication methods, such as a user name or login name.
User certificate
Click to browse for, and select, a user's certificate.
CA certificate
Click to browse for, and select, a Certificate Authority's certificate.
Private key
Click to browse for, and select, a user's private key file.
Private key password
Enter the user password corresponding to the user's private key.
8.3.9.1.2. Configuring Tunneled TLS Settings
Anonymous identity
This value is used as the unencrypted identity.
CA certificate
Click to browse for, and select, a Certificate Authority's certificate.
Inner authentication
PAP — Password Authentication Protocol.
MSCHAP — Challenge Handshake Authentication Protocol.
MSCHAPv2 — Microsoft Challenge Handshake Authentication Protocol version 2.
CHAP — Challenge Handshake Authentication Protocol.
Username
Enter the user name to be used in the authentication process.
Password
Enter the password to be used in the authentication process.
8.3.9.1.3. Configuring Protected EAP (PEAP) Settings
Anonymous Identity
This value is used as the unencrypted identity.
CA certificate
Click to browse for, and select, a Certificate Authority's certificate.
PEAP version
The version of Protected EAP to use. Automatic, 0 or 1.
Inner authentication
MSCHAPv2 — Microsoft Challenge Handshake Authentication Protocol version 2.
MD5 — Message Digest 5, a cryptographic hash function.
GTC — Generic Token Card.
Username
Enter the user name to be used in the authentication process.
Password
Enter the password to be used in the authentication process.

8.3.9.2. Configuring Wireless Security

Security
None — Do not encrypt the Wi-Fi connection.
WEP 40/128-bit Key — Wired Equivalent Privacy (WEP), from the IEEE 802.11 standard. Uses a single pre-shared key (PSK).
WEP 128-bit Passphrase — An MD5 hash of the passphrase will be used to derive a WEP key.
LEAP — Lightweight Extensible Authentication Protocol, from Cisco Systems.
Dynamic WEP (802.1x) — WEP keys are changed dynamically.
WPA & WPA2 Personal — Wi-Fi Protected Access (WPA), from the draft IEEE 802.11i standard. A replacement for WEP. Wi-Fi Protected Access II (WPA2), from the 802.11i-2004 standard. Personal mode uses a pre-shared key (WPA-PSK).
WPA & WPA2 Enterprise — WPA for use with a RADUIS authentication server to provide IEEE 802.1x network access control.
Password
Enter the password to be used in the authentication process.

Prevent Roaming On The Same Access Point

In the case of WPA and WPA2 (Personal and Enterprise), an option to select between Auto, WPA and WPA2 has been added. This option is intended for use with an access point that is offering both WPA and WPA2. Select one of the protocols if you would like to prevent roaming between the two protocols. Roaming between WPA and WPA2 on the same access point can cause loss of service.
Editing the Wireless Security tab and selecting the WPA protocol

Figure 8.16. Editing the Wireless Security tab and selecting the WPA protocol


8.3.9.3. Configuring PPP (Point-to-Point) Settings

Configure Methods
Use point-to-point encryption (MPPE)
Microsoft Point-To-Point Encryption protocol (RFC 3078).
Allow BSD data compression
PPP BSD Compression Protocol (RFC 1977).
Allow Deflate data compression
PPP Deflate Protocol (RFC 1979).
Use TCP header compression
Compressing TCP/IP Headers for Low-Speed Serial Links (RFC 1144).
Send PPP echo packets
LCP Echo-Request and Echo-Reply Codes for loopback tests (RFC 1661).

8.3.9.4. Configuring IPv4 Settings

Editing the IPv4 Settings Tab

Figure 8.17. Editing the IPv4 Settings Tab


The IPv4 Settings tab allows you to configure the method by which you connect to the Internet and enter IP address, route, and DNS information as required. The IPv4 Settings tab is available when you create and modify one of the following connection types: wired, wireless, mobile broadband, VPN or DSL.
If you are using DHCP to obtain a dynamic IP address from a DHCP server, you can simply set Method to Automatic (DHCP).
Setting the Method

Available IPv4 Methods by Connection Type

When you click the Method dropdown menu, depending on the type of connection you are configuring, you are able to select one of the following IPv4 connection methods. All of the methods are listed here according to which connection type or types they are associated with.
Method
Automatic (DHCP) — Choose this option if the network you are connecting to uses a DHCP server to assign IP addresses. You do not need to fill in the DHCP client ID field.
Automatic (DHCP) addresses only — Choose this option if the network you are connecting to uses a DHCP server to assign IP addresses but you want to assign DNS servers manually.
Link-Local Only — Choose this option if the network you are connecting to does not have a DHCP server and you do not want to assign IP addresses manually. Random addresses will be selected as per RFC 3927.
Shared to other computers — Choose this option if the interface you are configuring is for sharing an Internet or WAN connection.
Wired, Wireless and DSL Connection Methods
Manual — Choose this option if the network you are connecting to does not have a DHCP server and you want to assign IP addresses manually.
Mobile Broadband Connection Methods
Automatic (PPP) — Choose this option if the network you are connecting to uses a DHCP server to assign IP addresses.
Automatic (PPP) addresses only — Choose this option if the network you are connecting to uses a DHCP server to assign IP addresses but you want to assign DNS servers manually.
VPN Connection Methods
Automatic (VPN) — Choose this option if the network you are connecting to uses a DHCP server to assign IP addresses.
Automatic (VPN) addresses only — Choose this option if the network you are connecting to uses a DHCP server to assign IP addresses but you want to assign DNS servers manually.
DSL Connection Methods
Automatic (PPPoE) — Choose this option if the network you are connecting to uses a DHCP server to assign IP addresses.
Automatic (PPPoE) addresses only — Choose this option if the network you are connecting to uses a DHCP server to assign IP addresses but you want to assign DNS servers manually.
PPPoE Specific Configuration Steps
If more than one NIC is installed, and PPPoE will only be run over one NIC but not the other, then for correct PPPoE operation it is also necessary to lock the connection to the specific Ethernet device PPPoE is supposed to be run over. To lock the connection to one specific NIC, do one of the following:
  • Enter the MAC address in nm-connection-editor for that connection. Optionally select Connect automatically and Available to all users to make the connection come up without requiring user login after system start.
  • Set the hardware-address in the [802-3-ethernet] section in the appropriate file for that connection in /etc/NetworkManager/system-connections/ as follows:
    [802-3-ethernet]
    mac-address=00:11:22:33:44:55
    Mere presence of the file in /etc/NetworkManager/system-connections/ means that it is available to all users. Ensure that autoconnect=true appears in the [connection] section for the connection to be brought up without requiring user login after system start.
For information on configuring static routes for the network connection, go to Section 8.3.9.6, “Configuring Routes”.

8.3.9.5. Configuring IPv6 Settings

Method
Ignore — Choose this option if you want to disable IPv6 settings.
Automatic — Choose this option if the network you are connecting to uses a DHCP server to assign IP addresses.
Automatic, addresses only — Choose this option if the network you are connecting to uses a DHCP server to assign IP addresses but you want to assign DNS servers manually.
Manual — Choose this option if the network you are connecting to does not have a DHCP server and you want to assign IP addresses manually.
Link-Local Only — Choose this option if the network you are connecting to does not have a DHCP server and you do not want to assign IP addresses manually. Random addresses will be selected as per RFC 4862.
Shared to other computers — Choose this option if the interface you are configuring is for sharing an Internet or WAN connection.
Addresses
DNS servers — Enter a comma separated list of DNS servers.
Search domains — Enter a comma separated list of domain controllers.
For information on configuring static routes for the network connection, go to Section 8.3.9.6, “Configuring Routes”.

8.3.9.6. Configuring Routes

A host's routing table will be automatically populated with routes to directly connected networks. The routes are learned by observing the network interfaces when they are up. This section is for entering static routes to networks or hosts which can be reached by traversing an intermediate network or connection, such as a VPN or leased line.
Configuring static network routes

Figure 8.18. Configuring static network routes


Addresses
Address — The IP address of a network, sub-net or host.
Netmask — The netmask or prefix length of the IP address just entered.
Gateway — The IP address of the gateway leading to the network, sub-net or host.
Metric — A network cost, that is to say a preference value to give to this route. Lower values will be preferred over higher values.
Ignore automatically obtained routes
Select this check box to only use manually entered routes for this connection.
Use this connection only for resources on its network
Select this checkbox to prevent the connection from becoming the default route. Typical examples are where a connection is a VPN or a leased line to a head office and you do not want any Internet bound traffic to pass over the connection. Selecting this option means that only traffic specifically destined for routes learned automatically over the connection or entered here manually will be routed over the connection.

8.4. NetworkManager Architecture

Chapter 9. Network Interfaces

Under Red Hat Enterprise Linux, all network communications occur between configured software interfaces and physical networking devices connected to the system.
The configuration files for network interfaces are located in the /etc/sysconfig/network-scripts/ directory. The scripts used to activate and deactivate these network interfaces are also located here. Although the number and type of interface files can differ from system to system, there are three categories of files that exist in this directory:
  1. Interface configuration files
  2. Interface control scripts
  3. Network function files
The files in each of these categories work together to enable various network devices.
This chapter explores the relationship between these files and how they are used.

9.1. Network Configuration Files

Before delving into the interface configuration files, let us first itemize the primary configuration files used in network configuration. Understanding the role these files play in setting up the network stack can be helpful when customizing a Red Hat Enterprise Linux system.
The primary network configuration files are as follows:
/etc/hosts
The main purpose of this file is to resolve host names that cannot be resolved any other way. It can also be used to resolve host names on small networks with no DNS server. Regardless of the type of network the computer is on, this file should contain a line specifying the IP address of the loopback device (127.0.0.1) as localhost.localdomain. For more information, refer to the hosts(5) manual page.
/etc/resolv.conf
This file specifies the IP addresses of DNS servers and the search domain. Unless configured to do otherwise, the network initialization scripts populate this file. For more information about this file, refer to the resolv.conf(5) manual page.
/etc/sysconfig/network
This file specifies routing and host information for all network interfaces. It is used to contain directives which are to have global effect and not to be interface specific. For more information about this file and the directives it accepts, refer to Section D.1.13, “/etc/sysconfig/network”.
/etc/sysconfig/network-scripts/ifcfg-interface-name
For each network interface, there is a corresponding interface configuration script. Each of these files provide information specific to a particular network interface. Refer to Section 9.2, “Interface Configuration Files” for more information on this type of file and the directives it accepts.

Network interface names

Network interface names may be different on different hardware types. Refer to Appendix A, Consistent Network Device Naming for more information.

The /etc/sysconfig/networking/ directory

The /etc/sysconfig/networking/ directory is used by the now deprecated Network Administration Tool (system-config-network). Its contents should not be edited manually. Using only one method for network configuration is strongly encouraged, due to the risk of configuration deletion. For more information about configuring network interfaces using graphical configuration tools, refer to Chapter 8, NetworkManager.

9.2. Interface Configuration Files

Interface configuration files control the software interfaces for individual network devices. As the system boots, it uses these files to determine what interfaces to bring up and how to configure them. These files are usually named ifcfg-name, where name refers to the name of the device that the configuration file controls.

9.2.1. Ethernet Interfaces

One of the most common interface files is /etc/sysconfig/network-scripts/ifcfg-eth0, which controls the first Ethernet network interface card or NIC in the system. In a system with multiple NICs, there are multiple ifcfg-ethX files (where X is a unique number corresponding to a specific interface). Because each device has its own configuration file, an administrator can control how each interface functions individually.
The following is a sample ifcfg-eth0 file for a system using a fixed IP address:
DEVICE=eth0
BOOTPROTO=none
ONBOOT=yes
NETMASK=255.255.255.0
IPADDR=10.0.1.27
USERCTL=no
The values required in an interface configuration file can change based on other values. For example, the ifcfg-eth0 file for an interface using DHCP looks different because IP information is provided by the DHCP server:
DEVICE=eth0
BOOTPROTO=dhcp
ONBOOT=yes
NetworkManager is graphical configuration tool which provides an easy way to make changes to the various network interface configuration files (refer to Chapter 8, NetworkManager for detailed instructions on using this tool).
However, it is also possible to manually edit the configuration files for a given network interface.
Below is a listing of the configurable parameters in an Ethernet interface configuration file:
BONDING_OPTS=parameters
sets the configuration parameters for the bonding device, and is used in /etc/sysconfig/network-scripts/ifcfg-bondN (see Section 9.2.4, “Channel Bonding Interfaces”). These parameters are identical to those used for bonding devices in /sys/class/net/bonding_device/bonding, and the module parameters for the bonding driver as described in bonding Module Directives.
This configuration method is used so that multiple bonding devices can have different configurations. It is highly recommended to place all of your bonding options after the BONDING_OPTS directive in ifcfg-name. Do not specify options for the bonding device in /etc/modprobe.d/bonding.conf, or in the deprecated /etc/modprobe.conf file.
BOOTPROTO=protocol
where protocol is one of the following:
  • none — No boot-time protocol should be used.
  • bootp — The BOOTP protocol should be used.
  • dhcp — The DHCP protocol should be used.
BROADCAST=address
where address is the broadcast address. This directive is deprecated, as the value is calculated automatically with ipcalc.
DEVICE=name
where name is the name of the physical device (except for dynamically-allocated PPP devices where it is the logical name).
DHCP_HOSTNAME=name
where name is a short host name to be sent to the DHCP server. Use this option only if the DHCP server requires the client to specify a host name before receiving an IP address.
DHCPV6C=answer
where answer is one of the following:
  • yes — Use DHCP to obtain an IPv6 address for this interface.
  • no — Do not use DHCP to obtain an IPv6 address for this interface. This is the default value.
An IPv6 link-local address will still be assigned by default. The link-local address is based on the MAC address of the interface as per RFC 4862.
DHCPV6C_OPTIONS=answer
where answer is one of the following:
  • -P — Enable IPv6 prefix delegation.
  • -S — Use DHCP to obtain stateless configuration only, not addresses, for this interface.
  • -N — Restore normal operation after using the -T or -P options.
  • -T — Use DHCP to obtain a temporary IPv6 address for this interface.
  • -D — Override the default when selecting the type of DHCP Unique Identifier (DUID) to use.
    By default, the DHCPv6 client (dhclient) creates a DHCP Unique Identifier (DUID) based on the link-layer address (DUID-LL) if it is running in stateless mode (with the -S option, to not request an address), or it creates an identifier based on the link-layer address plus a timestamp (DUID-LLT) if it is running in stateful mode (without -S, requesting an address). The -D option overrides this default, with a value of either LL or LLT.
DNS{1,2}=address
where address is a name server address to be placed in /etc/resolv.conf if the PEERDNS directive is set to yes.
ETHTOOL_OPTS=options
where options are any device-specific options supported by ethtool. For example, if you wanted to force 100Mb, full duplex:
ETHTOOL_OPTS="autoneg off speed 100 duplex full"
Instead of a custom initscript, use ETHTOOL_OPTS to set the interface speed and duplex settings. Custom initscripts run outside of the network init script lead to unpredictable results during a post-boot network service restart.

Set autoneg off before changing speed or duplex settings

Changing speed or duplex settings almost always requires disabling auto-negotiation with the autoneg off option. This option needs to be stated first, as the option entries are order-dependent.
Refer to Section 9.7, “Ethtool” for more ethtool options.
HOTPLUG=answer
where answer is one of the following:
  • yes — This device should be activated when it is hot-plugged (this is the default option).
  • no — This device should not be activated when it is hot-plugged.
The HOTPLUG=no option can be used to prevent a channel bonding interface from being activated when a bonding kernel module is loaded.
Refer to Section 9.2.4, “Channel Bonding Interfaces” for more information about channel bonding interfaces.
HWADDR=MAC-address
where MAC-address is the hardware address of the Ethernet device in the form AA:BB:CC:DD:EE:FF. This directive must be used in machines containing more than one NIC to ensure that the interfaces are assigned the correct device names regardless of the configured load order for each NIC's module. This directive should not be used in conjunction with MACADDR.

Note

  • Persistent device names are now handled by /etc/udev/rules.d/70-persistent-net.rules.
  • HWADDR must not be used with System z network devices.
  • Refer to Section 25.3.3, "Mapping subchannels and network device names", in the Red Hat Enterprise Linux 6 Installation Guide.
IPADDR=address
where address is the IPv4 address.
IPV6ADDR=address
where address is the first static, or primary, IPv6 address on an interface.
The format is Address/Prefix-length. If no prefix length is specified, /64 is assumed. Note that this setting depends on IPV6INIT being enabled.
IPV6ADDR_SECONDARIES=address
where address is one or more, space separated, additional IPv6 addresses.
The format is Address/Prefix-length. If no prefix length is specified, /64 is assumed. Note that this setting depends on IPV6INIT being enabled.
IPV6INIT=answer
where answer is one of the following:
  • yes — Initialize this interface for IPv6 addressing.
  • no — Do not initialize this interface for IPv6 addressing. This is the default value.
    Note that this setting is required for IPv6 static, DHCP, or autoconf assignment of IPv6 addresses. An IPv6 link-local address will still be assigned by default. The link-local address is based on the MAC address of the interface as per RFC 4862.
    The global directive NETWORKING_IPV6 is required in the /etc/sysconfig/network conf file to globally enable IPv6 static, DHCP, or autoconf configuration. Refer to Section D.1.13, “/etc/sysconfig/network”
IPV6_AUTOCONF=answer
where answer is one of the following:
  • yes — Enable IPv6 autoconf configuration for this interface.
  • no — Disable IPv6 autoconf configuration for this interface.
If enabled, an IPv6 address will be requested using Neighbor Discovery (ND) from a router running the radvd daemon.
Note that the default value of IPV6_AUTOCONF depends on IPV6FORWARDING as follows:
  • If IPV6FORWARDING=yes, then IPV6_AUTOCONF will default to no.
  • If IPV6FORWARDING=no, then IPV6_AUTOCONF will default to yes and IPV6_ROUTER has no effect.
IPV6_MTU=value
where value is an optional dedicated MTU for this interface.
IPV6_PRIVACY=rfc3041
where rfc3041 optionally sets this interface to support RFC 3041 Privacy Extensions for Stateless Address Autoconfiguration in IPv6. Note that this setting depends on IPV6INIT option being enabled.
The default is for RFC 3041 support to be disabled. Stateless Autoconfiguration will derive addresses based on the MAC address, when available, using the modified EUI-64 method. The address is appended to a prefix but as the address is normally derived from the MAC address it is globally unique even when the prefix changes. In the case of a link-local address the prefix is fe80::/64 as per RFC 2462 IPv6 Stateless Address Autoconfiguration.
LINKDELAY=time
where time is the number of seconds to wait for link negotiation before configuring the device.
MACADDR=MAC-address
where MAC-address is the hardware address of the Ethernet device in the form AA:BB:CC:DD:EE:FF.
This directive is used to assign a MAC address to an interface, overriding the one assigned to the physical NIC. This directive should not be used in conjunction with the HWADDR directive.
MASTER=bond-interface
where bond-interface is the channel bonding interface to which the Ethernet interface is linked.
This directive is used in conjunction with the SLAVE directive.
Refer to Section 9.2.4, “Channel Bonding Interfaces” for more information about channel bonding interfaces.
NETMASK=mask
where mask is the netmask value.
NETWORK=address
where address is the network address. This directive is deprecated, as the value is calculated automatically with ipcalc.
NM_CONTROLLED=answer
where answer is one of the following:
  • yesNetworkManager is permitted to configure this device. This is the default behavior and can be omitted.
  • noNetworkManager is not permitted to configure this device.

Note

The NM_CONTROLLED directive is now, as of Red Hat Enterprise Linux 6.3, dependent on the NM_BOND_VLAN_ENABLED directive in /etc/sysconfig/network. If and only if that directive is present and is one of yes, y, or true, will NetworkManager detect and manage bonding and VLAN interfaces.
ONBOOT=answer
where answer is one of the following:
  • yes — This device should be activated at boot-time.
  • no — This device should not be activated at boot-time.
PEERDNS=answer
where answer is one of the following:
  • yes — Modify /etc/resolv.conf if the DNS directive is set. If using DHCP, then yes is the default.
  • no — Do not modify /etc/resolv.conf.
SLAVE=answer
where answer is one of the following:
  • yes — This device is controlled by the channel bonding interface specified in the MASTER directive.
  • no — This device is not controlled by the channel bonding interface specified in the MASTER directive.
This directive is used in conjunction with the MASTER directive.
Refer to Section 9.2.4, “Channel Bonding Interfaces” for more about channel bonding interfaces.
SRCADDR=address
where address is the specified source IP address for outgoing packets.
USERCTL=answer
where answer is one of the following:
  • yes — Non-root users are allowed to control this device.
  • no — Non-root users are not allowed to control this device.

9.2.2. Specific ifcfg Options for Linux on System z

SUBCHANNELS=<read_device_bus_id>, <write_device_bus_id>, <data_device_bus_id>
where <read_device_bus_id>, <write_device_bus_id>, and <data_device_bus_id> are the three device bus IDs representing a network device.
PORTNAME=myname;
where myname is the Open Systems Adapter (OSA) portname or LAN Channel Station (LCS) portnumber.
CTCPROT=answer
where answer is one of the following:
  • 0 — Compatibility mode, TCP/IP for Virtual Machines (used with non-Linux peers other than IBM S/390 and IBM System z operating systems). This is the default mode.
  • 1 — Extended mode, used for Linux-to-Linux Peers.
  • 3 — Compatibility mode for S/390 and IBM System z operating systems.
This directive is used in conjunction with the NETTYPE directive. It specifies the CTC protocol for NETTYPE='ctc'. The default is 0.
OPTION='answer'
where 'answer' is a quoted string of any valid sysfs attributes and their value. The Red Hat Enterprise Linux installer currently uses this to configure the layer mode, (layer2), and the relative port number, (portno), of QETH devices. For example:
OPTIONS='layer2=1 portno=0'

9.2.3. Required ifcfg Options for Linux on System z

NETTYPE=answer
where answer is one of the following:
  • ctc — Channel-to-Channel communication. For point-to-point TCP/IP or TTY.
  • lcs — LAN Channel Station (LCS).
  • qeth — QETH (QDIO Ethernet). This is the default network interface. It is the preferred installation method for supporting real or virtual OSA cards and HiperSockets devices.

9.2.4. Channel Bonding Interfaces

Red Hat Enterprise Linux allows administrators to bind multiple network interfaces together into a single channel using the bonding kernel module and a special network interface called a channel bonding interface. Channel bonding enables two or more network interfaces to act as one, simultaneously increasing the bandwidth and providing redundancy.
To create a channel bonding interface, create a file in the /etc/sysconfig/network-scripts/ directory called ifcfg-bondN, replacing N with the number for the interface, such as 0.
The contents of the file can be identical to whatever type of interface is getting bonded, such as an Ethernet interface. The only difference is that the DEVICE directive is bondN, replacing N with the number for the interface. The NM_CONTROLLED directive can be added to prevent NetworkManager from configuring this device.
The following is a sample channel bonding configuration file:

Example 9.1. Sample ifcfg-bond0 interface configuration file

DEVICE=bond0
IPADDR=192.168.1.1
NETMASK=255.255.255.0
ONBOOT=yes
BOOTPROTO=none
USERCTL=no
NM_CONTROLLED=no
BONDING_OPTS="bonding parameters separated by spaces"

After the channel bonding interface is created, the network interfaces to be bound together must be configured by adding the MASTER and SLAVE directives to their configuration files. The configuration files for each of the channel-bonded interfaces can be nearly identical.
For example, if two Ethernet interfaces are being channel bonded, both eth0 and eth1 may look like the following example:
DEVICE=ethN
BOOTPROTO=none
ONBOOT=yes
MASTER=bond0
SLAVE=yes
USERCTL=no
NM_CONTROLLED=no
In this example, replace N with the numerical value for the interface.
Support for bonding was added to NetworkManager in Red Hat Enterprise Linux 6.3. See Section 9.2.1, “Ethernet Interfaces” for an explanation of NM_CONTROLLED and the NM_BOND_VLAN_ENABLED directive.

Put all bonding module parameters in ifcfg-bondN files

Parameters for the bonding kernel module must be specified as a space-separated list in the BONDING_OPTS="bonding parameters" directive in the ifcfg-bondN interface file. Do not specify options for the bonding device in /etc/modprobe.d/bonding.conf, or in the deprecated /etc/modprobe.conf file. For further instructions and advice on configuring the bonding module and to view the list of bonding parameters, refer to Section 28.7.2, “Using Channel Bonding”.

9.2.5. Network Bridge

A network bridge is a Link Layer device which forwards traffic between networks based on MAC addresses and is therefore also referred to as a Layer 2 device. It makes forwarding decisions based on tables of MAC addresses which it builds by learning what hosts are connected to each network. A software bridge can be used within a Linux host in order to emulate a hardware bridge, for example in virtualization applications for sharing a NIC with one or more virtual NICs. This case will be illustrated here as an example.
To create a network bridge, create a file in the /etc/sysconfig/network-scripts/ directory called ifcfg-brN, replacing N with the number for the interface, such as 0.
The contents of the file is similar to whatever type of interface is getting bridged to, such as an Ethernet interface. The differences in this example are as follows:
  • The DEVICE directive is given an interface name as its argument in the format brN, where N is replaced with the number of the interface.
  • The TYPE directive is given an argument Bridge or Ethernet. This directive determines the device type and the argument is case sensitive.
  • The bridge interface configuration file now has the IP address and the physical interface has only a MAC address.
  • An extra directive, DELAY=0, is added to prevent the bridge from waiting while it monitors traffic, learns where hosts are located, and builds a table of MAC addresses on which to base its filtering decisions. The default delay of 30 seconds is not needed if no routing loops are possible.
  • The NM_CONTROLLED=no should be added to the Ethernet interface to prevent NetworkManager from altering the file. It can also be added to the bridge configuration file in case future versions of NetworkManager support bridge configuration.
The following is a sample bridge interface configuration file using a static IP address:

Example 9.2. Sample ifcfg-br0 interface configuration file

DEVICE=br0
TYPE=Bridge
IPADDR=192.168.1.1
NETMASK=255.255.255.0
ONBOOT=yes
BOOTPROTO=none
NM_CONTROLLED=no
DELAY=0

To complete the bridge another interface is created, or an existing interface is modified, and pointed to the bridge interface. The following is a sample Ethernet interface configuration file pointing to a bridge interface. Configure your physical interface in /etc/sysconfig/network-scripts/ifcfg-ethX, where X is a unique number corresponding to a specific interface, as follows:

Example 9.3. Sample ifcfg-ethX interface configuration file

DEVICE=ethX
TYPE=Ethernet
HWADDR=AA:BB:CC:DD:EE:FF
BOOTPROTO=none
ONBOOT=yes
NM_CONTROLLED=no
BRIDGE=br0

Note

For the DEVICE directive, almost any interface name could be used as it does not determine the device type. Other commonly used names include tap, dummy and bond for example. TYPE=Ethernet is not strictly required. If the TYPE directive is not set, the device is treated as an Ethernet device (unless it's name explicitly matches a different interface configuration file.)
You can refer to Section 9.2, “Interface Configuration Files” for a review of the directives and options used in network interface config files.

Warning

If you are configuring bridging on a remote host, and you are connected to that host over the physical NIC you are configuring, please consider the implications of losing connectivity before proceeding. You will lose connectivity when restarting the service and may not be able to regain connectivity if any errors have been made. Console, or out-of-band access is advised.
Restart the networking service, in order for the changes to take effect, as follows:
 service network restart 
An example of a network bridge formed from two or more bonded Ethernet interfaces will now be given as this is another common application in a virtualization environment. If you are not very familiar with the configuration files for bonded interfaces then please refer to Section 9.2.4, “Channel Bonding Interfaces”
Create or edit two or more Ethernet interface configuration files, which are to be bonded, as follows:
DEVICE=ethX
TYPE=Ethernet
USERCTL=no
SLAVE=yes
MASTER=bond0
BOOTPROTO=none
HWADDR=AA:BB:CC:DD:EE:FF
NM_CONTROLLED=no

Note

Using ethX as the interface name is common practice but almost any name could be used. Names such as tap, dummy and bond are commonly used.
Create or edit one interface configuration file, /etc/sysconfig/network-scripts/ifcfg-bond0, as follows:
DEVICE=bond0
ONBOOT=yes
BONDING_OPTS='mode=1 miimon=100'
BRIDGE=brbond0
NM_CONTROLLED=no
For further instructions and advice on configuring the bonding module and to view the list of bonding parameters, refer to Section 28.7.2, “Using Channel Bonding”.
Create or edit one interface configuration file, /etc/sysconfig/network-scripts/ifcfg-brbond0, as follows:
DEVICE=brbond0
ONBOOT=yes
TYPE=Bridge
IPADDR=192.168.1.1
NETMASK=255.255.255.0
NM_CONTROLLED=no
A network bridge consisting of two bonded Ethernet interfaces.

Figure 9.1. A network bridge consisting of two bonded Ethernet interfaces.


We now have two or more interface configuration files with the MASTER=bond0 directive. These point to the configuration file named /etc/sysconfig/network-scripts/ifcfg-bond0, which contains the DEVICE=bond0 directive. This ifcfg-bond0 in turn points to the /etc/sysconfig/network-scripts/ifcfg-brbond0 configuration file, which contains the IP address, and acts as an interface to the virtual networks inside the host.
Restart the networking service, in order for the changes to take effect, as follows:
 service network restart 

9.2.6. Setting Up 802.1q VLAN Tagging

  1. Ensure that the module is loaded by entering the following command:
     lsmod | grep 8021q
  2. If the module is not loaded, load it with the following command:
    modprobe 8021q
  3. Configure your physical interface in /etc/sysconfig/network-scripts/ifcfg-ethX, where X is a unique number corresponding to a specific interface, as follows:
    DEVICE=ethX
    TYPE=Ethernet
    BOOTPROTO=none
    ONBOOT=yes
  4. Configure the VLAN interface configuration in /etc/sysconfig/network-scripts. The configuration filename should be the physical interface plus a . character plus the VLAN ID number. For example, if the VLAN ID is 192, and the physical interface is eth0, then the configuration filename should be ifcfg-eth0.192:
    DEVICE=ethX.192
    BOOTPROTO=none
    ONBOOT=yes
    IPADDR=192.168.1.1
    NETMASK=255.255.255.0
    USERCTL=no
    NETWORK=192.168.1.0
    VLAN=yes
    If there is a need to configure a second VLAN, with for example, VLAN ID 193, on the same interface, eth0 , add a new file with the name eth0.193 with the VLAN configuration details.
  5. Restart the networking service, in order for the changes to take effect, as follows:
     service network restart 

9.2.7. Alias and Clone Files

Two lesser-used types of interface configuration files are alias and clone files. As the ip command of the iproute package now supports assigning multiple address to the same interface it is no longer necessary to use this method of binding multiple addresses to the same interface.

Note

At the time of writing, NetworkManager does not detect IP aliases in ifcfg files. For example, if ifcfg-eth0 and ifcfg-eth0:1 files are present, NetworkManager creates two connections, which will cause confusion.
For new installations, users should select the Manual method on the IPv4 or IPv6 tab in NetworkManager to assign multiple IP address to the same interface. For more information on using this tool, refer to Chapter 8, NetworkManager.
Alias interface configuration files, which are used to bind multiple addresses to a single interface, use the ifcfg-if-name:alias-value naming scheme.
For example, an ifcfg-eth0:0 file could be configured to specify DEVICE=eth0:0 and a static IP address of 10.0.0.2, serving as an alias of an Ethernet interface already configured to receive its IP information via DHCP in ifcfg-eth0. Under this configuration, eth0 is bound to a dynamic IP address, but the same physical network card can receive requests via the fixed, 10.0.0.2 IP address.

Warning

Alias interfaces do not support DHCP.
A clone interface configuration file should use the following naming convention: ifcfg-if-name-clone-name. While an alias file allows multiple addresses for an existing interface, a clone file is used to specify additional options for an interface. For example, a standard DHCP Ethernet interface called eth0, may look similar to this:
DEVICE=eth0
ONBOOT=yes
BOOTPROTO=dhcp
Since the default value for the USERCTL directive is no if it is not specified, users cannot bring this interface up and down. To give users the ability to control the interface, create a clone by copying ifcfg-eth0 to ifcfg-eth0-user and add the following line to ifcfg-eth0-user:
USERCTL=yes
This way a user can bring up the eth0 interface using the /sbin/ifup eth0-user command because the configuration options from ifcfg-eth0 and ifcfg-eth0-user are combined. While this is a very basic example, this method can be used with a variety of options and interfaces.
It is no longer possible to create alias and clone interface configuration files using a graphical tool. However, as explained at the beginning of this section, it is no longer necessary to use this method as it is now possible to directly assign multiple IP address to the same interface. For new installations, users should select the Manual method on the IPv4 or IPv6 tab in NetworkManager to assign multiple IP address to the same interface. For more information on using this tool, refer to Chapter 8, NetworkManager.

9.2.8. Dialup Interfaces

If you are connecting to the Internet via a dialup connection, a configuration file is necessary for the interface.
PPP interface files are named using the following format:
ifcfg-pppX
where X is a unique number corresponding to a specific interface.
The PPP interface configuration file is created automatically when wvdial, or Kppp is used to create a dialup account. It is also possible to create and edit this file manually.
The following is a typical /etc/sysconfig/network-scripts/ifcfg-ppp0 file:
DEVICE=ppp0
NAME=test
WVDIALSECT=test
MODEMPORT=/dev/modem
LINESPEED=115200
PAPNAME=test
USERCTL=true
ONBOOT=no
PERSIST=no
DEFROUTE=yes
PEERDNS=yes
DEMAND=no
IDLETIMEOUT=600
Serial Line Internet Protocol (SLIP) is another dialup interface, although it is used less frequently. SLIP files have interface configuration file names such as ifcfg-sl0.
Other options that may be used in these files include:
DEFROUTE=answer
where answer is one of the following:
  • yes — Set this interface as the default route.
  • no — Do not set this interface as the default route.
DEMAND=answer
where answer is one of the following:
  • yes — This interface allows pppd to initiate a connection when someone attempts to use it.
  • no — A connection must be manually established for this interface.
IDLETIMEOUT=value
where value is the number of seconds of idle activity before the interface disconnects itself.
INITSTRING=string
where string is the initialization string passed to the modem device. This option is primarily used in conjunction with SLIP interfaces.
LINESPEED=value
where value is the baud rate of the device. Possible standard values include 57600, 38400, 19200, and 9600.
MODEMPORT=device
where device is the name of the serial device that is used to establish the connection for the interface.
MTU=value
where value is the Maximum Transfer Unit (MTU) setting for the interface. The MTU refers to the largest number of bytes of data a frame can carry, not counting its header information. In some dialup situations, setting this to a value of 576 results in fewer packets dropped and a slight improvement to the throughput for a connection.
NAME=name
where name is the reference to the title given to a collection of dialup connection configurations.
PAPNAME=name
where name is the username given during the Password Authentication Protocol (PAP) exchange that occurs to allow connections to a remote system.
PERSIST=answer
where answer is one of the following:
  • yes — This interface should be kept active at all times, even if deactivated after a modem hang up.
  • no — This interface should not be kept active at all times.
REMIP=address
where address is the IP address of the remote system. This is usually left unspecified.
WVDIALSECT=name
where name associates this interface with a dialer configuration in /etc/wvdial.conf. This file contains the phone number to be dialed and other important information for the interface.

9.2.9. Other Interfaces

Other common interface configuration files include the following:
ifcfg-lo
A local loopback interface is often used in testing, as well as being used in a variety of applications that require an IP address pointing back to the same system. Any data sent to the loopback device is immediately returned to the host's network layer.

Do not manually edit the ifcfg-lo script

The loopback interface script, /etc/sysconfig/network-scripts/ifcfg-lo, should never be edited manually. Doing so can prevent the system from operating correctly.
ifcfg-irlan0
An infrared interface allows information between devices, such as a laptop and a printer, to flow over an infrared link. This works in a similar way to an Ethernet device except that it commonly occurs over a peer-to-peer connection.
ifcfg-plip0
A Parallel Line Interface Protocol (PLIP) connection works much the same way as an Ethernet device, except that it utilizes a parallel port.
Interface configuration files for Linux on System z include the following:
ifcfg-hsiN
A HiperSockets interface is an interface for high-speed TCP/IP communication within and across z/VM guest virtual machines and logical partitions (LPARs) on an IBM System z mainframe.

9.3. Interface Control Scripts

The interface control scripts activate and deactivate system interfaces. There are two primary interface control scripts that call on control scripts located in the /etc/sysconfig/network-scripts/ directory: /sbin/ifdown and /sbin/ifup.
The ifup and ifdown interface scripts are symbolic links to scripts in the /sbin/ directory. When either of these scripts are called, they require the value of the interface to be specified, such as:
ifup eth0

Use the ifup and ifdown interface scripts

The ifup and ifdown interface scripts are the only scripts that the user should use to bring up and take down network interfaces.
The following scripts are described for reference purposes only.
Two files used to perform a variety of network initialization tasks during the process of bringing up a network interface are /etc/rc.d/init.d/functions and /etc/sysconfig/network-scripts/network-functions. Refer to Section 9.6, “Network Function Files” for more information.
After verifying that an interface has been specified and that the user executing the request is allowed to control the interface, the correct script brings the interface up or down. The following are common interface control scripts found within the /etc/sysconfig/network-scripts/ directory:
ifup-aliases
Configures IP aliases from interface configuration files when more than one IP address is associated with an interface.
ifup-ippp and ifdown-ippp
Brings ISDN interfaces up and down.
ifup-ipv6 and ifdown-ipv6
Brings IPv6 interfaces up and down.
ifup-plip
Brings up a PLIP interface.
ifup-plusb
Brings up a USB interface for network connections.
ifup-post and ifdown-post
Contains commands to be executed after an interface is brought up or down.
ifup-ppp and ifdown-ppp
Brings a PPP interface up or down.
ifup-routes
Adds static routes for a device as its interface is brought up.
ifdown-sit and ifup-sit
Contains function calls related to bringing up and down an IPv6 tunnel within an IPv4 connection.
ifup-wireless
Brings up a wireless interface.

Be careful when removing or modifying network scripts!

Removing or modifying any scripts in the /etc/sysconfig/network-scripts/ directory can cause interface connections to act irregularly or fail. Only advanced users should modify scripts related to a network interface.
The easiest way to manipulate all network scripts simultaneously is to use the /sbin/service command on the network service (/etc/rc.d/init.d/network), as illustrated by the following command:
/sbin/service network action
Here, action can be either start, stop, or restart.
To view a list of configured devices and currently active network interfaces, use the following command:
/sbin/service network status

9.4. Static Routes and the Default Gateway

Static routes are for traffic that must not, or should not, go through the default gateway. Routing is usually handled by routing devices and therefore it is often not necessary to configure static routes on Red Hat Enterprise Linux servers or clients. Exceptions include traffic that must pass through an encrypted VPN tunnel or traffic that should take a less costly route. The default gateway is for any and all traffic which is not destined for the local network and for which no preferred route is specified in the routing table. The default gateway is traditionally a dedicated network router.

Static Routes

Use the ip route command to display the IP routing table. If static routes are required, they can be added to the routing table by means of the ip route add command and removed using the ip route del command. To add a static route to a host address, that is to say to a single IP address, issue the following command as root:
ip route add X.X.X.X
where X.X.X.X is the IP address of the host in dotted decimal notation. To add a static route to a network, that is to say to an IP address representing a range of IP addresses, issue the following command as root:
ip route add X.X.X.X/Y
where X.X.X.X is the IP address of the network in dotted decimal notation and Y is the network prefix. The network prefix is the number of enabled bits in the subnet mask. This format of network address slash prefix length is referred to as CIDR notation.
Static route configuration is stored per-interface in a /etc/sysconfig/network-scripts/route-interface file. For example, static routes for the eth0 interface would be stored in the /etc/sysconfig/network-scripts/route-eth0 file. The route-interface file has two formats: IP command arguments and network/netmask directives. These are described below.

The Default Gateway

The default gateway is determined by the network scripts which parse the /etc/sysconfig/network file first and then the network interface ifcfg files for interfaces that are up. The ifcfg files are parsed in numerically ascending order, and the last GATEWAY directive to be read is used to compose a default route in the routing table.
The default route can thus be indicated by means of the GATEWAY directive and can be specified either globally or in interface-specific configuration files. Specifying the gateway globally has certain advantages in static networking environments, especially if more than one network interface is present. It can make fault finding simpler if applied consistently. There is also the GATEWAYDEV directive, which is a global option. If multiple devices specify GATEWAY, and one interface uses the GATEWAYDEV directive, that directive will take precedence. This option is not recommend as it can have unexpected consequences if an interface goes down and it can complicate fault finding.
In dynamic network environments, where mobile hosts are managed by NetworkManager, gateway information is likely to be interface specific and is best left to be assigned by DHCP. In special cases where it is necessary to influence NetworkManager's selection of the exit interface to be used to reach a gateway, make use of the DEFROUTE=no command in the ifcfg files for those interfaces which do not lead to the default gateway.
Global default gateway configuration is stored in the /etc/sysconfig/network file. This file specifies gateway and host information for all network interfaces. For more information about this file and the directives it accepts, refer to Section D.1.13, “/etc/sysconfig/network”.

IP Command Arguments Format

If required in a per-interface configuration file, define a gateway on the first line. This is only required if the gateway is not set via DHCP and is not set globally as mentioned above:
default via X.X.X.X dev interface
X.X.X.X is the IP address of the default gateway. The interface is the interface that is connected to, or can reach, the default gateway. The dev option can be omitted, it is optional.
Define a static route. Each line is parsed as an individual route:
X.X.X.X/Y via X.X.X.X dev interface
X.X.X.X/Y is the network address and netmask for the static route. X.X.X.X and interface are the IP address and interface for the default gateway respectively. The X.X.X.X address does not have to be the default gateway IP address. In most cases, X.X.X.X will be an IP address in a different subnet, and interface will be the interface that is connected to, or can reach, that subnet. Add as many static routes as required.
The following is a sample route-eth0 file using the IP command arguments format. The default gateway is 192.168.0.1, interface eth0. The two static routes are for the 10.10.10.0/24 and 172.16.1.0/24 networks:
default via 192.168.0.1 dev eth0
10.10.10.0/24 via 192.168.0.1 dev eth0
172.16.1.0/24 via 192.168.0.1 dev eth0
Static routes should only be configured for other subnetworks. The above example is not necessary, since packets going to the 10.10.10.0/24 and 172.16.1.0/24 networks will use the default gateway anyway. Below is an example of setting static routes to a different subnet, on a machine in a 192.168.0.0/24 subnet. The example machine has an eth0 interface in the 192.168.0.0/24 subnet, and an eth1 interface (10.10.10.1) in the 10.10.10.0/24 subnet:
10.10.10.0/24 via 10.10.10.1 dev eth1
Specifying an exit interface is optional. It can be useful if you want to force traffic out of a specific interface. For example, in the case of a VPN, you can force traffic to a remote network to pass through a tun0 interface even when the interface is in a different sub-net to the destination network.

Duplicate default gateways

If the default gateway is already assigned from DHCP, the IP command arguments format can cause one of two errors during start-up, or when bringing up an interface from the down state using the ifup command: "RTNETLINK answers: File exists" or 'Error: either "to" is a duplicate, or "X.X.X.X" is a garbage.', where X.X.X.X is the gateway, or a different IP address. These errors can also occur if you have another route to another network using the default gateway. Both of these errors are safe to ignore.

Network/Netmask Directives Format

You can also use the network/netmask directives format for route-interface files. The following is a template for the network/netmask format, with instructions following afterwards:
 ADDRESS0=X.X.X.X NETMASK0=X.X.X.X GATEWAY0=X.X.X.X 
  • ADDRESS0=X.X.X.X is the network address for the static route.
  • NETMASK0=X.X.X.X is the netmask for the network address defined with ADDRESS0=X.X.X.X.
  • GATEWAY0=X.X.X.X is the default gateway, or an IP address that can be used to reach ADDRESS0=X.X.X.X
The following is a sample route-eth0 file using the network/netmask directives format. The default gateway is 192.168.0.1, interface eth0. The two static routes are for the 10.10.10.0/24 and 172.16.1.0/24 networks. However, as mentioned before, this example is not necessary as the 10.10.10.0/24 and 172.16.1.0/24 networks would use the default gateway anyway:
ADDRESS0=10.10.10.0
NETMASK0=255.255.255.0
GATEWAY0=192.168.0.1
ADDRESS1=172.16.1.0
NETMASK1=255.255.255.0
GATEWAY1=192.168.0.1
Subsequent static routes must be numbered sequentially, and must not skip any values. For example, ADDRESS0, ADDRESS1, ADDRESS2, and so on.
Below is an example of setting static routes to a different subnet, on a machine in the 192.168.0.0/24 subnet. The example machine has an eth0 interface in the 192.168.0.0/24 subnet, and an eth1 interface (10.10.10.1) in the 10.10.10.0/24 subnet:
ADDRESS0=10.10.10.0
NETMASK0=255.255.255.0
GATEWAY0=10.10.10.1
Note that if DHCP is used, it can assign these settings automatically.

9.5. Configuring IPv6 Tokenized Interface Identifiers

In a network, servers are generally given static addresses and these are usually configured manually to avoid relying on a DHCP server which may fail or run out of addresses. The IPv6 protocol introduced Stateless Address Autoconfiguration (SLAAC) which enables clients to assign themselves an address without relying on a DHCPv6 server. SLAAC derives the IPv6 address based on the interface hardware, therefore it should not be used for servers in case the hardware is changed and the associated SLAAC generated address changes with it. In an IPv6 environment, if the network prefix is changed, or the system is moved to a new location, any manually configured static addresses would have to be edited due to the changed prefix.
To address these problems, the IETF draft Tokenised IPv6 Identifiers has been implemented in the kernel together with corresponding additions to the ip utility. This enables the lower 64 bit interface identifier part of the IPv6 address to be based on a token, supplied by the administrator, leaving the network prefix, the higher 64 bits, to be obtained from router advertisements (RA). This means that if the network interface hardware is changed, the lower 64 bits of the address will not change, and if the system is moved to another network, the network prefix will be obtained from router advertisements automatically, thus no manual editing is required.
To configure an interface to use a tokenized IPv6 identifier, issue a command in the following format as root user:
~]# ip token set ::1a:2b:3c:4d/64 dev eth4
Where ::1a:2b:3c:4d/64 is the token to be used. This setting is not persistent. To make it persistent, add the command to an init script. See Section 9.3, “Interface Control Scripts”.
Using a memorable token is possible, but is limited to the range of valid hexadecimal digits. For example, for a DNS server, which traditionally uses port 53, a token of ::53/64 could be used.
To view all the configured IPv6 tokens, issue the following command:
~]$ ip token
     token :: dev eth0 
     token :: dev eth1 
     token :: dev eth2 
     token :: dev eth3 
     token ::1a:2b:3c:4d dev eth4
To view the configured IPv6 token for a specific interface, issue the following command:
~]$ ip token get dev eth4
     token ::1a:2b:3c:4d dev eth4
Note that adding a token to an interface will replace a previously allocated token, and in turn invalidate the address derived from it. Supplying a new token causes a new address to be generated and applied, but this process will leave any other addresses unchanged. In other words, a new tokenized identifier only replaces a previously existing tokenized identifier, not any other IP address.

Note

Take care not to add the same token to more than one system or interface as the duplicate address detection (DAD) mechanism will not be able to resolve the problem. Once a token is set, it cannot be cleared or reset, except by rebooting the machine.

9.6. Network Function Files

Red Hat Enterprise Linux makes use of several files that contain important common functions used to bring interfaces up and down. Rather than forcing each interface control file to contain these functions, they are grouped together in a few files that are called upon when necessary.
The /etc/sysconfig/network-scripts/network-functions file contains the most commonly used IPv4 functions, which are useful to many interface control scripts. These functions include contacting running programs that have requested information about changes in the status of an interface, setting host names, finding a gateway device, verifying whether or not a particular device is down, and adding a default route.
As the functions required for IPv6 interfaces are different from IPv4 interfaces, a /etc/sysconfig/network-scripts/network-functions-ipv6 file exists specifically to hold this information. The functions in this file configure and delete static IPv6 routes, create and remove tunnels, add and remove IPv6 addresses to an interface, and test for the existence of an IPv6 address on an interface.

9.7. Ethtool

Ethtool is a utility for configuration of Network Interface Cards (NICs). This utility allows querying and changing settings such as speed, port, auto-negotiation, PCI locations and checksum offload on many network devices, especially Ethernet devices.
We present here a short selection of often used ethtool commands together with some useful commands that are not well known. For a full list of commands type ethtool -h or refer to the man page, ethtool(8), for a more comprehensive list and explanation. The first two examples are information queries and show the use of the different formats of the command.
But first, the command structure:
ethtool [option...] devname
where option is none or more options, and devname is your Network Interface Card (NIC). For example eth0 or em1.
ethtool
The ethtool command with only a device name as an option is used to print the current settings of the specified device. It takes the following form:
ethtool devname
where devname is your NIC. For example eth0 or em1.
Some values can only be obtained when the command is run as root. Here is an example of the output when the command is run as root:
~]# ethtool em1
Settings for em1:
	Supported ports: [ TP ]
	Supported link modes:   10baseT/Half 10baseT/Full
	                        100baseT/Half 100baseT/Full
	                        1000baseT/Full
	Supported pause frame use: No
	Supports auto-negotiation: Yes
	Advertised link modes:  10baseT/Half 10baseT/Full
	                        100baseT/Half 100baseT/Full
	                        1000baseT/Full
	Advertised pause frame use: No
	Advertised auto-negotiation: Yes
	Speed: 1000Mb/s
	Duplex: Full
	Port: Twisted Pair
	PHYAD: 2
	Transceiver: internal
	Auto-negotiation: on
	MDI-X: on
	Supports Wake-on: pumbg
	Wake-on: g
	Current message level: 0x00000007 (7)
			       drv probe link
	Link detected: yes
Issue the following command, using the short or long form of the argument, to query the specified network device for associated driver information:
ethtool -i, --driver devname
where devname is your Network Interface Card (NIC). For example eth0 or em1.
Here is an example of the output:
~]$ ethtool -i em1
driver: e1000e
version: 2.0.0-k
firmware-version: 0.13-3
bus-info: 0000:00:19.0
supports-statistics: yes
supports-test: yes
supports-eeprom-access: yes
supports-register-dump: yes
Here follows a list of command options to query, identify or reset the device. They are in the usual -short and --long form:
--statistics
The --statistics or -S queries the specified network device for NIC and driver statistics. It takes the following form:
-S, --statistics devname
where devname is your NIC.
--identify
The --identify or -p option initiates adapter-specific action intended to enable an operator to easily identify the adapter by sight. Typically this involves blinking one or more LEDs on the specified network port. It takes the following form:
-p, --identify devname integer
where integer is length of time in seconds to perform the action,
and devname is your NIC.
--show-time-stamping
The --show-time-stamping or -T option queries the specified network device for time stamping parameters. It takes the following form:
-T, --show-time-stamping devname
where devname is your NIC.
--show-offload
The --show-features, or --show-offload, or -k option queries the specified network device for the state of protocol offload and other features. It takes the following form:
-k, --show-features, --show-offload devname
where devname is your NIC.
--test
The --test or -t option is used to perform tests on a Network Interface Card. It takes the following form:
-t, --test word devname
where word is one of the following:
  • offline — Perform a comprehensive set of tests. Service will be interrupted.
  • online — Perform a reduced set of tests. Service should not be interrupted.
  • external_lb — Perform full set of tests including loopback tests while fitted with a loopback cable.
and devname is your NIC.
Changing some or all settings of the specified network device requires the -s or --change option. All the following options are only applied if the -s or --change option is also specified. For the sake of clarity we will omit it here.
To make these settings permanent you can make use of the ETHTOOL_OPTS directive. It can be used in interface configuration files to set the desired options when the network interface is brought up. Refer to Section 9.2.1, “Ethernet Interfaces” for more details on how to use this directive.
--offload
The --features, or --offload, or -K option changes the offload parameters and other features of the specified network device. It takes the following form:
-K, --features, --offload devname feature boolean
where feature is a built-in or kernel supplied feature,
boolean is one of ON or OFF,
and devname is your NIC.
The ethtool(8) man page lists most features. As the feature set is dependent on the NIC driver, you should consult the driver documentation for features not listed in the man page.
--speed
The --speed option is used to set the speed in megabits per second (Mb/s). Omitting the speed value will show the supported device speeds. It takes the following form:
--speed number devname
where number is the speed in megabits per second (Mb/s),
and devname is your NIC.
--duplex
The --duplex option is used to set the transmit and receive mode of operation. It takes the following form:
 --duplex word devname
where word is one of the following:
  • half — Sets half-duplex mode. Usually used when connected to a hub.
  • full — Sets full-duplex mode. Usually used when connected to a switch or another host.
and devname is your NIC.
--port
The --port option is used to select the device port . It takes the following form:
--port value devname
where value is one of the following:
  • tp — An Ethernet interface using Twisted-Pair cable as the medium.
  • aui — Attachment Unit Interface (AUI). Normally used with hubs.
  • bnc — An Ethernet interface using BNC connectors and co-axial cable.
  • mii — An Ethernet interface using a Media Independent Interface (MII).
  • fibre — An Ethernet interface using Optical Fibre as the medium.
and devname is your NIC.
--autoneg
The --autoneg option is used to control auto-negotiation of network speed and mode of operation (full-duplex or half-duplex mode). If auto-negotiation is enabled you can initiate re-negotiation of network speeds and mode of operation by using the -r, --negotiate option. You can display the auto-negotiation state using the --a, --show-pause option.
It takes the following form:
--autoneg value devname
where value is one of the following:
  • yes — Allow auto-negotiating of network speed and mode of operation.
  • no — Do not allow auto-negotiating of network speed and mode of operation.
and devname is your NIC.
--advertise
The --advertise option is used to set what speeds and modes of operation (duplex mode) are advertised for auto-negotiation. The argument is one or more hexadecimal values from Table 9.1, “Ethtool advertise options: speed and mode of operation”.
It takes the following form:
--advertise option devname
where option is one or more of the hexadecimal values from the table below and devname is your NIC.

Table 9.1. Ethtool advertise options: speed and mode of operation

Hex Value Speed Duplex Mode IEEE standard?
0x001 10
Half
Yes
0x002 10
Full
Yes
0x004 100
Half
Yes
0x008 100
Full
Yes
0x010 1000 Half No
0x020 1000
Full
Yes
0x8000 2500 Full Yes
0x1000 10000 Full Yes
0x20000 20000MLD2 Full No
0x20000 20000MLD2 Full No
0x40000 20000KR2 Full No

--phyad
The --phyad option is used to change the physical address. Often referred to as the MAC or hardware address but in this context referred to as the physical address.
It takes the following form:
--phyad physical_address devname
where physical_address is the physical address in hexadecimal format and devname is your NIC.
--xcvr
The --xcvr option is used to select the transceiver type. Currently only internal and external can be specified. In the future other types might be added.
It takes the following form:
--xcvr word devname
where word is one of the following:
  • internal — Use internal transceiver.
  • external — Use external transceiver.
and devname is your NIC.
--wol
The --wol option is used to set Wake-on-LAN options. Not all devices support this. The argument to this option is a string of characters specifying which options to enable.
It takes the following form:
--wol value devname
where value is one or more of the following:
  • p — Wake on PHY activity.
  • u — Wake on unicast messages.
  • m — Wake on multicast messages.
  • b — Wake on broadcast messages.
  • g — Wake-on-Lan; wake on receipt of a "magic packet".
  • s — Enable security function using password for Wake-on-Lan.
  • d — Disable Wake-on-Lan and clear all settings.
and devname is your NIC.
--sopass
The --sopass option is used to set the SecureOn password. The argument to this option must be 6 bytes in Ethernet MAC hexadecimal format (xx:yy:zz:aa:bb:cc).
It takes the following form:
--sopass xx:yy:zz:aa:bb:cc devname
where xx:yy:zz:aa:bb:cc is the password in the same format as a MAC address and devname is your NIC.
--msglvl
The --msglvl option is used to set the driver message-type flags by name or number. The precise meanings of these type flags differ between drivers.
It takes the following form:
--msglvl message_type devname
where message_type is one of:
  • message type name in plain text.
  • hexadecimal number indicating the message type.
and devname is your NIC.
The defined message type names and numbers are shown in the table below:

Table 9.2. Driver message type

Message Type Hex Value Description
drv 0x0001
General driver status
probe 0x0002
Hardware probing
link 0x0004
Link state
timer 0x0008
Periodic status check
ifdown 0x0010 Interface being brought down
ifup 0x0020
Interface being brought up
rx_err 0x0040 Receive error
tx_err 0x0080 Transmit error
intr 0x0200 Interrupt handling
tx_done 0x0400 Transmit completion
rx_status 0x0800 Receive completion
pktdata 0x1000 Packet contents
hw 0x2000 Hardware status
wol 0x4000 Wake-on-LAN status

9.8. Additional Resources

The following are resources which explain more about network interfaces.

9.8.1. Installed Documentation

/usr/share/doc/initscripts-version/sysconfig.txt
A guide to available options for network configuration files, including IPv6 options not covered in this chapter.

9.8.2. Useful Websites

http://linux-ip.net/gl/ip-cref/
This document contains a wealth of information about the ip command, which can be used to manipulate routing tables, among other things. The information can also be found in the ip-cref.ps file by installing the iproute-doc sub-package from the optional content channel.

Chapter 10. Configure SCTP

10.1. Introduction to Streaming Control Transport Protocol (SCTP)

Streaming Control Transport Protocol (SCTP) was proposed in RFC 2960, since made obsolete by RFC 4960. It is a message oriented, reliable transport protocol with direct support for multihoming that runs on top of the Internet Protocol (IPv4 and IPv6). Many IETF RFCs for SCTP have been published since the original proposal.
SCTP was originally developed to overcome some of the limitations of TCP for use in call control signaling. For example:
  • Resistance to malicious attacks, such as TCP SYN flood, was the top priority.
  • Path level redundancy against link failures was required.
  • A message-oriented rather than byte-oriented protocol was desired.
  • TCP is reliable and has order-of-transmission delivery, but in telecommunication signaling data can be out of order.

Table 10.1. Compare SCTP To TCP and UDP

Service or Characteristic SCTP TCP UDP
Packet header size 12 bytes 20–60 bytes 8 bytes
Transport-layer packet Datagram Segment Datagram
Connection oriented Yes Yes No
Reliable or Unreliable transport Both Reliable Unreliable
Partial Reliability Optional No No
Ordered or Unordered delivery Both Ordered Unordered
Preserve message boundary Yes No Yes
Congestion control Yes Yes No
ECN support Yes Yes No
Path MTU discovery Yes Yes No
Application PDU fragmentation Yes Yes No
Application PDU bundling Yes Yes No
Multi-homing support Yes No No
Multi-streaming support Yes No No
SYN Flood Protection Yes No N/A
Keep-alive heartbeat support Yes No No

10.1.1. Comparison of TCP and SCTP Handshaking

TCP uses a three-way handshake. The client sends a synchronize (SYN) packet, the server responds with a synchronize-acknowledgment (SYN-ACK) packet, and then the client confirms the connection with an acknowledgment (ACK) packet. Unfortunately, this makes the server vulnerable to a SYN-flood resource starvation denial-of-service attack if many connection attempts are initiated but not acknowledged.
SCTP has a four-way handshake that includes the passing of a cookie. The client sends an initiation (INIT) packet, the server responds with an initiation-acknowledgment (INIT-ACK) packet which includes a state cookie. The client returns the state cookie (COOKIE-ECHO), and then the server confirms the connection with a cookie-acknowledgment (COOKIE-ACK) packet. An attacker, typically using a spoofed address, would not get the state cookie. SCTP does not reserve any memory until it receives and authenticates the returned state cookie, thus limiting the potential for this type of attack.
Comparison of TCP and SCTP Handshaking

Figure 10.1. Comparison of TCP and SCTP Handshaking


10.2. Understanding SCTP

SCTP was initially designed to carry Signaling System 7 (SS7) signals over IP, but it has since been found useful in other applications. The original motivation for SCTP, or for putting SS7 on IP networks, was because the infrastructure was available and the hardware and maintenance costs were lower than traditional telecommunication equipment. SCTP is more than just the speed of UDP with some of the reliability of TCP, it also provides: Multihoming, bundled multiple streams, partial reliability, message boundaries, and application inspection and control.
It is worth noting that SCTP is the transport protocol for the SIGTRAN protocols, which are an extension of the SS7 protocol family. The Internet Engineering Task Force (IETF) defined layer 2, 3, and 4 protocols that are compatible with SS7:
  • Message Transfer Part (MTP) level 2 (M2UA and M2PA)
  • Message Transfer Part (MTP) level 3 (M3UA)
  • Message Transfer Part (MTP) level 2 (M2UA and M2PA)
These use a SCTP transport mechanism. This suite of protocols is called SIGTRAN.
When SCTP connects to a remote peer, it exchanges what is called an INIT chunk with its new peer. In this INIT chunk is an initial list of the IP addresses the originating host may use for communicating with the peer. That tells the remote host that traffic on this connection may originate from any of the IP addresses listed. Addresses may be added or deleted later using the ASCONF chunk. SCTP establishes associations between one of the IP addresses at each endpoint. Multiple streams of messages then traverse each association. Multiple associations may be active at the same time. If an association fails, and the application allows it, the streams on that association migrate to another association.

10.2.1. Bundled Streams

An association is in effect, a pair of communicating IP addresses within a connection. The streams are actually just ordered sequences of data chunks, messages, within an association. Each data chunk has what is called a stream identifier, sid, associated with it. Applications, within a connection, can create multiple streams (by default up to 65k), each of which has its own ordering rules. By default, a stream is always delivered in order (this is the TCP-like nature of SCTP). But applications can configure connections so that certain packets can be dropped if they do not reach their destination in a pre-configured amount of time.

10.2.2. Partial Reliability

TCP is mandated reliability, that is to say, if you send data on a TCP socket, it waits in that socket until it gets an acknowledgment indicating that it was received by its peer. By default, SCTP is exactly the same. However, an application may indicate, programmatically, that a given message can be discarded if it is not delivered within a preset time limit. Such a discard can then result in notifications to the application on both ends of the connection about lost data. Note that there is also the ability to send data unordered using the SCTP_UNORDERED flag, which removes all ordering constraints for that message within a stream. See man page sctp_sendmsg(3) (available as part of the development package).

10.2.3. Message Boundary Preservation

TCP is stream oriented, that is to say, there is a loss of information when sending data because the receiver has no idea in what quanta you sent the data. For example, a sender might write 2 bytes, then 4 bytes, then 6 bytes to a socket. The receiver might call the read function and get 12 bytes all at once or they might get 6 bytes on one read and 6 bytes on another read. In TCP there is no guarantee on the amount of data read, and normally no preservation of record boundaries. However, you could disable Nagle's algorithm (using TCP_NODELAY), as it is done in interactive SSH sessions. This comes at the expense of efficiency because there is a trade-off in the ratio of packet meta-data to the actual data payload. In addition, on slow links many packets can be in flight, potentially leading to congestion if packet size is very small.
SCTP is different, it can operate in two modes, sequential packet, and stream. Both modes provide for record boundary preservation. In sequential packet mode, it operates similar to UDP in that each packet is an atomic record, and if the sender sends an 804 byte datagram, the receiver will eventually get a single 804 byte datagram at the other end. It will be ordered within its stream, just like TCP , but it will be handled as an atomic unit, and not delivered to the application in pieces. In stream mode, on the other hand, data is sent as in TCP , in various bits and pieces, with no attention paid to record boundaries. However, you can still determine the boundaries of a record, because SCTP will insert the MSG_EOR flag at the end of a datagram sent from the sender side. In stream mode you may only get a partial record, but if your program pays attention to the MSG_EOR flag in the msghdr structure, you can tell when your multiple reads reach the end of a record. If there is the end of one message and the start of a second message in the queue, the read function will return only the end of one message with the MSG_EOR flag set. This allows your application to distinguish where one message ends and the next begins. See man pages sctp(7) and sctp_recvmsg(3) for details on the flags mentioned.

10.2.4. Protocol Event Notifications

SCTP provides information, in the form of notifications, about the protocol behavior but it is up to the application to make use of the information. Applications have to be written to use and manage the features that SCTP provides. The notifications are defined in the SCTP socket extensions: http://tools.ietf.org/html/draft-ietf-tsvwg-sctpsocket-26. The document defines all the notifications that SCTP can send to a user, as well as all the control messages that an application can send to the protocol. The API defined by this document is implemented by the lksctp user space library. It is important to note, that all of the messages defined in this API are handled in-band on a socket, so users interested in programming with SCTP are encouraged to use this library rather than reimplement the whole document within their application. The API provides wrapper functions around the traditional BSD socket calls, so the user can use this library instead of the standard POSIX API for SCTP sockets.

10.3. When To Use SCTP

SCTP should be considered for use when:
  • There are sufficient traffic levels to justify the overhead of association establishment, and congestion and flow control measures.
  • There is a requirement for framing of reliable data streams.
  • There is a transfer of multiple independent message sequences that are unrelated.
  • There is a transfer of messages that do not need to be delivered in sequence.
  • There is a requirement for network layer redundancy.

10.4. When Not To Use SCTP

SCTP should not be considered for use when:
  • There are small amounts of unrelated transactions generated and sent towards a destination.
  • There is an orientation towards byte-stream as opposed to message transfer.

10.5. Compare SCTP to Bonding and Network Teaming

Bonding is a link layer activity, that is to say, bonds are created and managed at the link layer whereas multihoming in SCTP takes place at the network layer and is therefore a routing operation. By selecting new source addresses on the host to send from, you cause different routes to be looked up, and different network paths are taken to reach a remote host.
It is also important to note that detecting a link layer failure is a very rapid operation. Routing operations however, require failure detection times that are orders of magnitude larger (due to the increased network path latencies, and need for retransmissions to ensure that a failure has actually occurred). This is the reason multihoming is not an acceptable replacement for bonding.

Table 10.2. Compare Bonding Teaming and SCTP

Characteristic Bonding Teaming SCTP
Monitoring layer Link layer Link layer Network layer
Failover time sub second intervals sub second intervals multiple seconds
Multihome support No No Yes
User-space runtime control Limited Full Full
Extensibility Hard Easy Yes

10.6. Using SCTP

Support for SCTP is implemented in the kernel and netfilter has support for SCTP provided by the xt_sctp module. The lksctp-tools project provides a Linux user space library, libsctp, for SCTP including C language header files, netinet/sctp.h, for accessing SCTP specific application programming interfaces not provided by the standard sockets.
In order to make use of SCTP, a programmer must write an application, and for this purpose the lksctp-tools-devel packages are available. The package also includes a number of extra man pages.
SCTP is also supported for IP Virtual Server (IPVS), also known as Linux Virtual Server (LVS).

10.6.1. Check if SCTP is installed

To check if lksctp-tools is installed, enter the following command as root:
~]# yum install lksctp-tools

10.6.2. Install SCTP

SCTP is implemented in the kernel. The lksctp-tools project provides a Linux user space library, libsctp, for SCTP including C language header files (netinet/sctp.h) for accessing SCTP specific application programming interfaces not provided by the standard sockets.
To check if lksctp-tools is installed, enter the following command as root:
~]# yum install lksctp-tools
For development work, in programming a new application for SCTP, the development packages are also required. To install them, issue the following command as root.
~]# yum install lksctp-tools-devel
This will also install a number of extra man pages listed in the sctp(7) man page.

10.6.3. Configuring SCTP

Since SCTP runs on top of IP, it does not need to be configured directly, the administrator only needs to do IP configuration (addresses, default gateway, nameservers, and so on). Everything else is handled programmatically by the application using SCTP. That is to say, the application configures its link(s) for things like multihoming, failover, and reliability. Binding to multiple addresses in SCTP potentially allows for multiple paths to a peer and enables the use of the SCTP multihoming feature.
SCTP configuration is very heavily biased towards the application developer and programmatic configuration. Using SCTP requires a good understanding of IP and routing. Any settings relevant to IP are relevant to SCTP as well. SCTP, like TCP and UDP, is IPv6 aware, so any settings relevant to IPv6 are also applicable.

10.6.4. Tune SCTP

The administrative interface for SCTP is the sysctl settings listed in the proc/sys/net/sctp/* Variables section of /usr/share/doc/kernel-doc-kernel_version/Documentation/networking/ip-sysctl.txt.
The developers of lksctp-tools have made an effort to default the settings listed in ip-sysctl.txt to the most sensible value for general use. So when in doubt, just use the default settings. Beyond that, specific settings reflect a specific workload or function. For instance, you might set rto_min n to a lower value to ensure that data is retransmitted more quickly to ensure a faster recovery from failure, but you do so at the risk of possibly retransmitting data chunk needlessly, thereby creating more network traffic, and slowing overall throughput.
The Red Hat Knowledgebase article Troubleshooting Stream Control Transmission Protocol (SCTP) contains further useful information.

10.6.5. Troubleshooting SCTP

The lksctp-tools includes a utility sctp_darn when can be used for basic testing. The command formats are as follows:
sctp_darn -H localhost -P localport [-h remotehost] [-p remoteport] -l, --listen
This will print messages received from the peer. Alternatively:
sctp_darn -H localhost -P localport [-h remotehost] [-p remoteport] --s, --send
will send messages to the peer. For more options enter the command sctp_darn and a list of options will be displayed.
It is important to remember the programmatic control aspect of SCTP. Therefore much of how SCTP is used stems from the application telling the protocol how to act. Therefore detailed knowledge of the custom application using SCTP is important in drawing up a test plan and a test case to reproduce the problem. A test case should be given to all those involved in testing. With knowledge of the application and test cases, detailed network diagrams should be used to obtain and examine copies of routing tables and tcpdump data from nodes likely to be carrying the SCTP traffic. Note that tcpdump needs to be used on all relevant interfaces and often on interfaces that do not seem to be relevant. So it is important to check that the tcpdump data includes all the links for which you have routes on the system. The SCTP dissector in Wireshark works well and is well maintained and can therefore be of great help in examining SCTP traffic.

10.7. Additional Resources

The following sources of information provide additional resources regarding SCTP and lksctp-tools.

10.7.1. Installed Documentation

  • sctp(7) man page — Describes mapping of SCTP into a sockets API.
  • /usr/share/doc/kernel-doc-kernel_version/Documentation/networking/ip-sysctl.txt — The section proc/sys/net/sctp/* Variables contains a list of options for SCTP.

10.7.2. Useful Websites

http://lksctp.sourceforge.net/
Technical information about the SCTP tools, includes information on supported specifications.
http://www.rfc-editor.org/info/rfc2960
RFC 2960 is the original proposal for the SCTP specification.
http://www.rfc-editor.org/info/rfc3286
An Introduction to the Stream Control Transmission Protocol (SCTP).
http://www.rfc-editor.org/info/rfc4960
RFC 4960 is the revised proposal for SCTP which obsoletes RFC 2960.
http://www.csm.ornl.gov/~dunigan/net100/sctp.html
A comparison between TCP and SCTP including state machine diagrams.
http://www.ibm.com/developerworks/library/l-sctp/
A good overview of SCTP including useful diagrams.
https://access.redhat.com/site/articles/219653
A Red Hat Knowledgebase article Troubleshooting Stream Control Transmission Protocol (SCTP).
https://access.redhat.com/site/solutions/58453
A Red Hat Knowledgebase solution A multi-homing example program for SCTP.

Part IV. Infrastructure Services

This part provides information how to configure services and daemons, configure authentication, and enable remote logins.

Table of Contents

11. Services and Daemons
11.1. Configuring the Default Runlevel
11.2. Configuring the Services
11.2.1. Using the Service Configuration Utility
11.2.2. Using the ntsysv Utility
11.2.3. Using the chkconfig Utility
11.3. Running Services
11.3.1. Determining the Service Status
11.3.2. Starting a Service
11.3.3. Stopping a Service
11.3.4. Restarting a Service
11.4. Additional Resources
11.4.1. Installed Documentation
11.4.2. Related Books
12. Configuring Authentication
12.1. Configuring System Authentication
12.1.1. Launching the Authentication Configuration Tool UI
12.1.2. Selecting the Identity Store for Authentication
12.1.3. Configuring Alternative Authentication Features
12.1.4. Configuring Authentication from the Command Line
12.1.5. Using Custom Home Directories
12.2. Using and Caching Credentials with SSSD
12.2.1. About SSSD
12.2.2. Setting up the sssd.conf File
12.2.3. Starting and Stopping SSSD
12.2.4. SSSD and System Services
12.2.5. Configuring Services: NSS
12.2.6. Configuring Services: PAM
12.2.7. Configuring Services: autofs
12.2.8. Configuring Services: sudo
12.2.9. Configuring Services: OpenSSH and Cached Keys
12.2.10. SSSD and Identity Providers (Domains)
12.2.11. Creating Domains: LDAP
12.2.12. Creating Domains: Identity Management (IdM)
12.2.13. Creating Domains: Active Directory
12.2.14. Configuring Domains: Active Directory as an LDAP Provider (Alternative)
12.2.15. Domain Options: Setting Username Formats
12.2.16. Domain Options: Enabling Offline Authentication
12.2.17. Domain Options: Setting Password Expirations
12.2.18. Domain Options: Using DNS Service Discovery
12.2.19. Domain Options: Using IP Addresses in Certificate Subject Names (LDAP Only)
12.2.20. Creating Domains: Proxy
12.2.21. Creating Domains: Kerberos Authentication
12.2.22. Creating Domains: Access Control
12.2.23. Creating Domains: Primary Server and Backup Servers
12.2.24. Installing SSSD Utilities
12.2.25. SSSD and UID and GID Numbers
12.2.26. Creating Local System Users
12.2.27. Seeding Users into the SSSD Cache During Kickstart
12.2.28. Managing the SSSD Cache
12.2.29. Downgrading SSSD
12.2.30. Using NSCD with SSSD
12.2.31. Troubleshooting SSSD
13. OpenSSH
13.1. The SSH Protocol
13.1.1. Why Use SSH?
13.1.2. Main Features
13.1.3. Protocol Versions
13.1.4. Event Sequence of an SSH Connection
13.2. Configuring OpenSSH
13.2.1. Configuration Files
13.2.2. Starting an OpenSSH Server
13.2.3. Requiring SSH for Remote Connections
13.2.4. Using a Key-Based Authentication
13.3. OpenSSH Clients
13.3.1. Using the ssh Utility
13.3.2. Using the scp Utility
13.3.3. Using the sftp Utility
13.4. More Than a Secure Shell
13.4.1. X11 Forwarding
13.4.2. Port Forwarding
13.5. Additional Resources
13.5.1. Installed Documentation
13.5.2. Useful Websites

Chapter 11. Services and Daemons

Maintaining security on your system is extremely important, and one approach for this task is to manage access to system services carefully. Your system may need to provide open access to particular services (for example, httpd if you are running a web server). However, if you do not need to provide a service, you should turn it off to minimize your exposure to possible bug exploits.
This chapter explains the concept of runlevels, and describes how to set the default one. It also covers the setup of the services to be run in each of these runlevels, and provides information on how to start, stop, and restart the services on the command line using the service command.

Keep the system secure

When you allow access for new services, always remember that both the firewall and SELinux need to be configured as well. One of the most common mistakes committed when configuring a new service is neglecting to implement the necessary firewall configuration and SELinux policies to allow access for it. For more information, refer to the Red Hat Enterprise Linux 6 Security Guide.

11.1. Configuring the Default Runlevel

A runlevel is a state, or mode, defined by services that are meant to be run when this runlevel is selected. Seven numbered runlevels exist (indexed from 0):

Table 11.1. Runlevels in Red Hat Enterprise Linux

Runlevel Description
0 Used to halt the system. This runlevel is reserved and cannot be changed.
1 Used to run in a single-user mode. This runlevel is reserved and cannot be changed.
2 Not used by default. You are free to define it yourself.
3 Used to run in a full multi-user mode with a command line user interface.
4 Not used by default. You are free to define it yourself.
5 Used to run in a full multi-user mode with a graphical user interface.
6 Used to reboot the system. This runlevel is reserved and cannot be changed.

To check in which runlevel you are operating, type the following:
~]$ runlevel
N 5
The runlevel command displays previous and current runlevel. In this case it is number 5, which means the system is running in a full multi-user mode with a graphical user interface.
The default runlevel can be changed by modifying the /etc/inittab file, which contains a line near the end of the file similar to the following:
id:5:initdefault:
To do so, edit this file as root and change the number on this line to the desired value. The change will take effect the next time you reboot the system.

11.2. Configuring the Services

To allow you to configure which services are started at boot time, Red Hat Enterprise Linux is shipped with the following utilities: the Service Configuration graphical application, the ntsysv text user interface, and the chkconfig command line tool.

Enabling the irqbalance service

To ensure optimal performance on POWER architecture, it is recommended that the irqbalance service is enabled. In most cases, this service is installed and configured to run during the Red Hat Enterprise Linux 6 installation. To verify that irqbalance is running, as root, type the following at a shell prompt:
~]# service irqbalance status
irqbalance (pid  1234) is running...
For information on how to enable and run a service using a graphical user interface, refer to Section 11.2.1, “Using the Service Configuration Utility”. For instructions on how to perform these task on the command line, see Section 11.2.3, “Using the chkconfig Utility” and Section 11.3, “Running Services” respectively.

11.2.1. Using the Service Configuration Utility

The Service Configuration utility is a graphical application developed by Red Hat to configure which services are started in a particular runlevel, as well as to start, stop, and restart them from the menu. To start the utility, select SystemAdministrationServices from the panel, or type the command system-config-services at a shell prompt.
The Service Configuration utility

Figure 11.1. The Service Configuration utility


The utility displays the list of all available services (services from the /etc/rc.d/init.d/ directory, as well as services controlled by xinetd) along with their description and the current status. For a complete list of used icons and an explanation of their meaning, see Table 11.2, “Possible service states”.
Note that unless you are already authenticated, you will be prompted to enter the superuser password the first time you make a change.

Table 11.2. Possible service states

Icon Description
The service is enabled.
The service is disabled.
The service is enabled for selected runlevels only.
The service is running.
The service is stopped.
There is something wrong with the service.
The status of the service is unknown.

11.2.1.1. Enabling and Disabling a Service

To enable a service, select it from the list and either click the Enable button on the toolbar, or choose ServiceEnable from the main menu.
To disable a service, select it from the list and either click the Disable button on the toolbar, or choose ServiceDisable from the main menu.

11.2.1.2. Starting, Restarting, and Stopping a Service

To start a service, select it from the list and either click the Start button on the toolbar, or choose ServiceStart from the main menu. Note that this option is not available for services controlled by xinetd, as they are started by it on demand.
To restart a running service, select it from the list and either click the Restart button on the toolbar, or choose ServiceRestart from the main menu. Note that this option is not available for services controlled by xinetd, as they are started and stopped by it automatically.
To stop a service, select it from the list and either click the Stop button on the toolbar, or choose ServiceStop from the main menu. Note that this option is not available for services controlled by xinetd, as they are stopped by it when their job is finished.

11.2.1.3. Selecting Runlevels

To enable the service for certain runlevels only, select it from the list and either click the Customize button on the toolbar, or choose ServiceCustomize from the main menu. Then select the checkbox beside each runlevel in which you want the service to run. Note that this option is not available for services controlled by xinetd.

11.2.2. Using the ntsysv Utility

The ntsysv utility is a command line application with a simple text user interface to configure which services are to be started in selected runlevels. To start the utility, type ntsysv at a shell prompt as root.
The ntsysv utility

Figure 11.2. The ntsysv utility


The utility displays the list of available services (the services from the /etc/rc.d/init.d/ directory) along with their current status and a description obtainable by pressing F1. For a list of used symbols and an explanation of their meaning, see Table 11.3, “Possible service states”.

Table 11.3. Possible service states

Symbol Description
[*] The service is enabled.
[ ] The service is disabled.

11.2.2.1. Enabling and Disabling a Service

To enable a service, navigate through the list using the Up and Down arrows keys, and select it with the Spacebar. An asterisk (*) appears in the brackets.
To disable a service, navigate through the list using the Up and Down arrows keys, and toggle its status with the Spacebar. An asterisk (*) in the brackets disappears.
Once you are done, use the Tab key to navigate to the Ok button, and confirm the changes by pressing Enter. Keep in mind that ntsysv does not actually start or stop the service. If you need to start or stop the service immediately, use the service command as described in Section 11.3.2, “Starting a Service”.

11.2.2.2. Selecting Runlevels

By default, the ntsysv utility only affects the current runlevel. To enable or disable services for other runlevels, as root, run the command with the additional --level option followed by numbers from 0 to 6 representing each runlevel you want to configure:
ntsysv --level runlevels
For example, to configure runlevels 3 and 5, type:
~]# ntsysv --level 35

11.2.3. Using the chkconfig Utility

The chkconfig utility is a command line tool that allows you to specify in which runlevel to start a selected service, as well as to list all available services along with their current setting. Note that with the exception of listing, you must have superuser privileges to use this command.

11.2.3.1. Listing the Services

To display a list of system services (services from the /etc/rc.d/init.d/ directory, as well as the services controlled by xinetd), either type chkconfig --list, or use chkconfig with no additional arguments. You will be presented with an output similar to the following:
~]# chkconfig --list
NetworkManager  0:off   1:off   2:on    3:on    4:on    5:on    6:off
abrtd           0:off   1:off   2:off   3:on    4:off   5:on    6:off
acpid           0:off   1:off   2:on    3:on    4:on    5:on    6:off
anamon          0:off   1:off   2:off   3:off   4:off   5:off   6:off
atd             0:off   1:off   2:off   3:on    4:on    5:on    6:off
auditd          0:off   1:off   2:on    3:on    4:on    5:on    6:off
avahi-daemon    0:off   1:off   2:off   3:on    4:on    5:on    6:off
... several lines omitted ...
wpa_supplicant  0:off   1:off   2:off   3:off   4:off   5:off   6:off

xinetd based services:
        chargen-dgram:  off
        chargen-stream: off
        cvs:            off
        daytime-dgram:  off
        daytime-stream: off
        discard-dgram:  off
... several lines omitted ...
        time-stream:    off
Each line consists of the name of the service followed by its status (on or off) for each of the seven numbered runlevels. For example, in the listing above, NetworkManager is enabled in runlevel 2, 3, 4, and 5, while abrtd runs in runlevel 3 and 5. The xinetd based services are listed at the end, being either on, or off.
To display the current settings for a selected service only, use chkconfig --list followed by the name of the service:
chkconfig --list service_name
For example, to display the current settings for the sshd service, type:
~]# chkconfig --list sshd
sshd            0:off   1:off   2:on    3:on    4:on    5:on    6:off
You can also use this command to display the status of a service that is managed by xinetd. In that case, the output will only contain the information whether the service is enabled or disabled:
~]# chkconfig --list rsync
rsync           off

11.2.3.2. Enabling a Service

To enable a service in runlevels 2, 3, 4, and 5, type the following at a shell prompt as root:
chkconfig service_name on
For example, to enable the httpd service in these four runlevels, type:
~]# chkconfig httpd on
To enable a service in certain runlevels only, add the --level option followed by numbers from 0 to 6 representing each runlevel in which you want the service to run:
chkconfig service_name on --level runlevels
For instance, to enable the abrtd service in runlevels 3 and 5, type:
~]# chkconfig abrtd on --level 35
The service will be started the next time you enter one of these runlevels. If you need to start the service immediately, use the service command as described in Section 11.3.2, “Starting a Service”.
Do not use the --level option when working with a service that is managed by xinetd, as it is not supported. For example, to enable the rsync service, type:
~]# chkconfig rsync on
If the xinetd daemon is running, the service is immediately enabled without having to manually restart the daemon.

11.2.3.3. Disabling a Service

To disable a service in runlevels 2, 3, 4, and 5, type the following at a shell prompt as root:
chkconfig service_name off
For instance, to disable the httpd service in these four runlevels, type:
~]# chkconfig httpd off
To disable a service in certain runlevels only, add the --level option followed by numbers from 0 to 6 representing each runlevel in which you do not want the service to run:
chkconfig service_name off --level runlevels
For instance, to disable the abrtd in runlevels 2 and 4, type:
~]# chkconfig abrtd off --level 24
The service will be stopped the next time you enter one of these runlevels. If you need to stop the service immediately, use the service command as described in Section 11.3.3, “Stopping a Service”.
Do not use the --level option when working with a service that is managed by xinetd, as it is not supported. For example, to disable the rsync service, type:
~]# chkconfig rsync off
If the xinetd daemon is running, the service is immediately disabled without having to manually restart the daemon.

11.3. Running Services

The service utility allows you to start, stop, or restart the services from the /etc/init.d/ directory.

11.3.1. Determining the Service Status

To determine the current status of a service, type the following at a shell prompt:
service service_name status
For example, to determine the status of the httpd service, type:
~]# service httpd status
httpd (pid  7474) is running...
To display the status of all available services at once, run the service command with the --status-all option:
~]# service --status-all
abrt (pid  1492) is running...
acpid (pid  1305) is running...
atd (pid  1540) is running...
auditd (pid  1103) is running...
automount (pid 1315) is running...
Avahi daemon is running
cpuspeed is stopped
... several lines omitted ...
wpa_supplicant (pid  1227) is running...
Note that you can also use the Service Configuration utility as described in Section 11.2.1, “Using the Service Configuration Utility”.

11.3.2. Starting a Service

To start a service, type the following at a shell prompt as root:
service service_name start
For example, to start the httpd service, type:
~]# service httpd start
Starting httpd:                                            [  OK  ]

11.3.3. Stopping a Service

To stop a running service, type the following at a shell prompt as root:
service service_name stop
For example, to stop the httpd service, type:
~]# service httpd stop
Stopping httpd:                                            [  OK  ]

11.3.4. Restarting a Service

To restart the service, type the following at a shell prompt as root:
service service_name restart
For example, to restart the httpd service, type:
~]# service httpd restart
Stopping httpd:                                            [  OK  ]
Starting httpd:                                            [  OK  ]

11.4. Additional Resources

11.4.1. Installed Documentation

  • chkconfig(8) — a manual page for the chkconfig utility.
  • ntsysv(8) — a manual page for the ntsysv utility.
  • service(8) — a manual page for the service utility.
  • system-config-services(8) — a manual page for the system-config-services utility.

11.4.2. Related Books

Red Hat Enterprise Linux 6 Security Guide
A guide to securing Red Hat Enterprise Linux 6. It contains valuable information on how to set up the firewall, as well as the configuration of SELinux.

Chapter 12. Configuring Authentication

12.1. Configuring System Authentication
12.1.1. Launching the Authentication Configuration Tool UI
12.1.2. Selecting the Identity Store for Authentication
12.1.3. Configuring Alternative Authentication Features
12.1.4. Configuring Authentication from the Command Line
12.1.5. Using Custom Home Directories
12.2. Using and Caching Credentials with SSSD
12.2.1. About SSSD
12.2.2. Setting up the sssd.conf File
12.2.3. Starting and Stopping SSSD
12.2.4. SSSD and System Services
12.2.5. Configuring Services: NSS
12.2.6. Configuring Services: PAM
12.2.7. Configuring Services: autofs
12.2.8. Configuring Services: sudo
12.2.9. Configuring Services: OpenSSH and Cached Keys
12.2.10. SSSD and Identity Providers (Domains)
12.2.11. Creating Domains: LDAP
12.2.12. Creating Domains: Identity Management (IdM)
12.2.13. Creating Domains: Active Directory
12.2.14. Configuring Domains: Active Directory as an LDAP Provider (Alternative)
12.2.15. Domain Options: Setting Username Formats
12.2.16. Domain Options: Enabling Offline Authentication
12.2.17. Domain Options: Setting Password Expirations
12.2.18. Domain Options: Using DNS Service Discovery
12.2.19. Domain Options: Using IP Addresses in Certificate Subject Names (LDAP Only)
12.2.20. Creating Domains: Proxy
12.2.21. Creating Domains: Kerberos Authentication
12.2.22. Creating Domains: Access Control
12.2.23. Creating Domains: Primary Server and Backup Servers
12.2.24. Installing SSSD Utilities
12.2.25. SSSD and UID and GID Numbers
12.2.26. Creating Local System Users
12.2.27. Seeding Users into the SSSD Cache During Kickstart
12.2.28. Managing the SSSD Cache
12.2.29. Downgrading SSSD
12.2.30. Using NSCD with SSSD
12.2.31. Troubleshooting SSSD
Authentication is the way that a user is identified and verified to a system. The authentication process requires presenting some sort of identity and credentials, like a username and password. The credentials are then compared to information stored in some data store on the system. In Red Hat Enterprise Linux, the Authentication Configuration Tool helps configure what kind of data store to use for user credentials, such as LDAP.
For convenience and potentially part of single sign-on, Red Hat Enterprise Linux can use a central daemon to store user credentials for a number of different data stores. The System Security Services Daemon (SSSD) can interact with LDAP, Kerberos, and external applications to verify user credentials. The Authentication Configuration Tool can configure SSSD along with NIS, Winbind, and LDAP, so that authentication processing and caching can be combined.

12.1. Configuring System Authentication

When a user logs into a Red Hat Enterprise Linux system, that user presents some sort of credential to establish the user identity. The system then checks those credentials against the configured authentication service. If the credentials match and the user account is active, then the user is authenticated. (Once a user is authenticated, then the information is passed to the access control service to determine what the user is permitted to do. Those are the resources the user is authorized to access.)
The information to verify the user can be located on the local system or the local system can reference a user database on a remote system, such as LDAP or Kerberos.
The system must have a configured list of valid account databases for it to check for user authentication. On Red Hat Enterprise Linux, the Authentication Configuration Tool has both GUI and command-line options to configure any user data stores.
A local system can use a variety of different data stores for user information, including Lightweight Directory Access Protocol (LDAP), Network Information Service (NIS), and Winbind. Additionally, both LDAP and NIS data stores can use Kerberos to authenticate users.

Important

If a medium or high security level is set during installation or with the Security Level Configuration Tool, then the firewall prevents NIS authentication. For more information about firewalls, see the "Firewalls" section of the Security Guide.

12.1.1. Launching the Authentication Configuration Tool UI

  1. Log into the system as root.
  2. Open the System.
  3. Select the Administration menu.
  4. Select the Authentication item.
Alternatively, run the system-config-authentication command.

Important

Any changes take effect immediately when the Authentication Configuration Tool UI is closed.
There are two configuration tabs in the Authentication dialog box:
  • Identity & Authentication, which configures the resource used as the identity store (the data repository where the user IDs and corresponding credentials are stored).
  • Advanced Options, which allows authentication methods other than passwords or certificates, like smart cards and fingerprint.

12.1.2. Selecting the Identity Store for Authentication

The Identity & Authentication tab sets how users should be authenticated. The default is to use local system authentication, meaning the users and their passwords are checked against local system accounts. A Red Hat Enterprise Linux machine can also use external resources which contain the users and credentials, including LDAP, NIS, and Winbind.
Local Authentication

Figure 12.1. Local Authentication


12.1.2.1. Configuring LDAP Authentication

Either the openldap-clients package or the sssd package is used to configure an LDAP server for the user database. Both packages are installed by default.
  1. Open the Authentication Configuration Tool, as in Section 12.1.1, “Launching the Authentication Configuration Tool UI”.
  2. Select LDAP in the User Account Database drop-down menu.
  3. Set the information that is required to connect to the LDAP server.
    • LDAP Search Base DN gives the root suffix or distinguished name (DN) for the user directory. All of the user entries used for identity/authentication will exist below this parent entry. For example, ou=people,dc=example,dc=com.
      This field is optional. If it is not specified, then the System Security Services Daemon (SSSD) attempts to detect the search base using the namingContexts and defaultNamingContext attributes in the LDAP server's configuration entry.
    • LDAP Server gives the URL of the LDAP server. This usually requires both the hostname and port number of the LDAP server, such as ldap://ldap.example.com:389.
      Entering the secure protocol in the URL, ldaps://, enables the Download CA Certificate button.
    • Use TLS to encrypt connections sets whether to use Start TLS to encrypt the connections to the LDAP server. This enables a secure connection over a standard port.
      Selecting TLS enables the Download CA Certificate button, which retrieves the issuing CA certificate for the LDAP server from whatever certificate authority issued it. The CA certificate must be in the privacy enhanced mail (PEM) format.

      Important

      Do not select Use TLS to encrypt connections if the server URL uses a secure protocol (ldaps). This option uses Start TLS, which initiates a secure connection over a standard port; if a secure port is specified, then a protocol like SSL must be used instead of Start TLS.
  4. Select the authentication method. LDAP allows simple password authentication or Kerberos authentication.
    The LDAP password option uses PAM applications to use LDAP authentication. This option requires either a secure (ldaps://) URL or the TLS option to connect to the LDAP server.

12.1.2.2. Configuring NIS Authentication

  1. Install the ypbind package. This is required for NIS services, but is not installed by default.
    [root@server ~]# yum install ypbind
    When the ypbind service is installed, the portmap and ypbind services are started and enabled to start at boot time.
  2. Open the Authentication Configuration Tool, as in Section 12.1.1, “Launching the Authentication Configuration Tool UI”.
  3. Select NIS in the User Account Database drop-down menu.
  4. Set the information to connect to the NIS server, meaning the NIS domain name and the server hostname. If the NIS server is not specified, the authconfig daemon scans for the NIS server.
  5. Select the authentication method. NIS allows simple password authentication or Kerberos authentication.
For more information about NIS, see the "Securing NIS" section of the Security Guide.

12.1.2.3. Configuring Winbind Authentication

  1. Install the samba-winbind package. This is required for Windows integration features in Samba services, but is not installed by default.
    [root@server ~]# yum install samba-winbind
  2. Open the Authentication Configuration Tool, as in Section 12.1.1, “Launching the Authentication Configuration Tool UI”.
  3. Select Winbind in the User Account Database drop-down menu.
  4. Set the information that is required to connect to the Microsoft Active Directory domain controller.
    • Winbind Domain gives the Windows domain to connect to.
      This should be in the Windows 2000 format, such as DOMAIN.
    • Security Model sets the security model to use for Samba clients. authconfig supports four types of security models:
      • ads configures Samba to act as a domain member in an Active Directory Server realm. To operate in this mode, the krb5-server package must be installed and Kerberos must be configured properly.
      • domain has Samba validate the username/password by authenticating it through a Windows primary or backup domain controller, much like a Windows server.
      • server has a local Samba server validate the username/password by authenticating it through another server, such as a Windows server. If the server authentication attempt fails, the system then attempts to authenticate using user mode.
      • user requires a client to log in with a valid username and password. This mode does support encrypted passwords.
        The username format must be domain\user, such as EXAMPLE\jsmith.

        Note

        When verifying that a given user exists in the Windows domain, always use Windows 2000-style formats and escape the backslash (\) character. For example:
        [root@server ~]# getent passwd domain\\user DOMAIN\user:*:16777216:16777216:Name Surname:/home/DOMAIN/user:/bin/bash
        This is the default option.
    • Winbind ADS Realm gives the Active Directory realm that the Samba server will join. This is only used with the ads security model.
    • Winbind Domain Controllers gives the domain controller to use. For more information about domain controllers, refer to Section 19.1.6.3, “Domain Controller”.
    • Template Shell sets which login shell to use for Windows user account settings.
    • Allow offline login allows authentication information to be stored in a local cache. The cache is referenced when a user attempts to authenticate to system resources while the system is offline.
For more information about the winbindd service, refer to Section 19.1.2, “Samba Daemons and Related Services”.

12.1.2.4. Using Kerberos with LDAP or NIS Authentication

Both LDAP and NIS authentication stores support Kerberos authentication methods. Using Kerberos has a couple of benefits:
  • It uses a security layer for communication while still allowing connections over standard ports.
  • It automatically uses credentials caching with SSSD, which allows offline logins.
Using Kerberos authentication requires the krb5-libs and krb5-workstation packages.
The Kerberos password option from the Authentication Method drop-down menu automatically opens the fields required to connect to the Kerberos realm.
Kerberos Fields

Figure 12.2. Kerberos Fields


  • Realm gives the name for the realm for the Kerberos server. The realm is the network that uses Kerberos, composed of one or more key distribution centers (KDC) and a potentially large number of clients.
  • KDCs gives a comma-separated list of servers that issue Kerberos tickets.
  • Admin Servers gives a list of administration servers running the kadmind process in the realm.
  • Optionally, use DNS to resolve server hostname and to find additional KDCs within the realm.
For more information about Kerberos, refer to section "Using Kerberos" of the Red Hat Enterprise Linux 6 Managing Single Sign-On and Smart Cards guide.

12.1.3. Configuring Alternative Authentication Features

The Authentication Configuration Tool also configures settings related to authentication behavior, apart from the identity store. This includes entirely different authentication methods (fingerprint scans and smart cards) or local authentication rules. These alternative authentication options are configured in the Advanced Options tab.
Advanced Options

Figure 12.3. Advanced Options


12.1.3.1. Using Fingerprint Authentication

When there is appropriate hardware available, the Enable fingerprint reader support option allows fingerprint scans to be used to authenticate local users in addition to other credentials.

12.1.3.2. Setting Local Authentication Parameters

There are two options in the Local Authentication Options area which define authentication behavior on the local system:
  • Enable local access control instructs the /etc/security/access.conf file to check for local user authorization rules.
  • Password Hashing Algorithm sets the hashing algorithm to use to encrypt locally-stored passwords.

12.1.3.3. Enabling Smart Card Authentication

When there are appropriate smart card readers available, a system can accept smart cards (or tokens) instead of other user credentials to authenticate.
Once the Enable smart card support option is selected, then the behaviors of smart card authentication can be defined:
  • Card Removal Action tells the system how to respond when the card is removed from the card reader during an active session. A system can either ignore the removal and allow the user to access resources as normal, or a system can immediately lock until the smart card is supplied.
  • Require smart card login sets whether a smart card is required for logins or simply allowed for logins. When this option is selected, all other methods of authentication are immediately blocked.

    Warning

    Do not select this option until you have successfully authenticated to the system using a smart card.
Using smart cards requires the pam_pkcs11 package.

12.1.3.4. Creating User Home Directories

There is an option (Create home directories on the first login) to create a home directory automatically the first time that a user logs in.
This option is beneficial with accounts that are managed centrally, such as with LDAP. However, this option should not be selected if a system like automount is used to manage user home directories.

12.1.4. Configuring Authentication from the Command Line

The authconfig command-line tool updates all of the configuration files and services required for system authentication, according to the settings passed to the script. Along with allowing all of the identity and authentication configuration options that can be set through the UI, the authconfig tool can also be used to create backup and kickstart files.
For a complete list of authconfig options, check the help output and the man page.

12.1.4.1. Tips for Using authconfig

There are some things to remember when running authconfig:
  • With every command, use either the --update or --test option. One of those options is required for the command to run successfully. Using --update writes the configuration changes. --test prints the changes to stdout but does not apply the changes to the configuration.
  • Each enable option has a corresponding disable option.

12.1.4.2. Configuring LDAP User Stores

To use an LDAP identity store, use the --enableldap. To use LDAP as the authentication source, use --enableldapauth and then the requisite connection information, like the LDAP server name, base DN for the user suffix, and (optionally) whether to use TLS. The authconfig command also has options to enable or disable RFC 2307bis schema for user entries, which is not possible through the Authentication Configuration UI.
Be sure to use the full LDAP URL, including the protocol (ldap or ldaps) and the port number. Do not use a secure LDAP URL (ldaps) with the --enableldaptls option.
authconfig --enableldap --enableldapauth --ldapserver=ldap://ldap.example.com:389,ldap://ldap2.example.com:389 --ldapbasedn="ou=people,dc=example,dc=com" --enableldaptls --ldaploadcacert=https://ca.server.example.com/caCert.crt --update
Instead of using --ldapauth for LDAP password authentication, it is possible to use Kerberos with the LDAP user store. These options are described in Section 12.1.4.5, “Configuring Kerberos Authentication”.

12.1.4.3. Configuring NIS User Stores

To use a NIS identity store, use the --enablenis. This automatically uses NIS authentication, unless the Kerberos parameters are explicitly set, so it uses Kerberos authentication (Section 12.1.4.5, “Configuring Kerberos Authentication”). The only parameters are to identify the NIS server and NIS domain; if these are not used, then the authconfig service scans the network for NIS servers.
authconfig --enablenis --nisdomain=EXAMPLE --nisserver=nis.example.com --update

12.1.4.4. Configuring Winbind User Stores

Windows domains have several different security models, and the security model used in the domain determines the authentication configuration for the local system.
For user and server security models, the Winbind configuration requires only the domain (or workgroup) name and the domain controller hostnames.
authconfig --enablewinbind --enablewinbindauth --smbsecurity=user|server  --enablewinbindoffline --smbservers=ad.example.com --smbworkgroup=EXAMPLE --update

Note

The username format must be domain\user, such as EXAMPLE\jsmith.
When verifying that a given user exists in the Windows domain, always use Windows 2000-style formats and escape the backslash (\) character. For example:
[root@server ~]# getent passwd domain\\user DOMAIN\user:*:16777216:16777216:Name Surname:/home/DOMAIN/user:/bin/bash
For ads and domain security models, the Winbind configuration allows additional configuration for the template shell and realm (ads only). For example:
authconfig --enablewinbind --enablewinbindauth --smbsecurity ads  --enablewinbindoffline --smbservers=ad.example.com --smbworkgroup=EXAMPLE --smbrealm EXAMPLE.COM --winbindtemplateshell=/bin/sh --update
There are a lot of other options for configuring Windows-based authentication and the information for Windows user accounts, such as name formats, whether to require the domain name with the username, and UID ranges. These options are listed in the authconfig help.

12.1.4.5. Configuring Kerberos Authentication

Both LDAP and NIS allow Kerberos authentication to be used in place of their native authentication mechanisms. At a minimum, using Kerberos authentication requires specifying the realm, the KDC, and the administrative server. There are also options to use DNS to resolve client names and to find additional admin servers.
authconfig NIS or LDAP options --enablekrb5 --krb5realm EXAMPLE --krb5kdc kdc.example.com:88,server.example.com:88 --krb5adminserver server.example.com:749 --enablekrb5kdcdns --enablekrb5realmdns --update

12.1.4.6. Configuring Local Authentication Settings

The Authentication Configuration Tool can also control some user settings that relate to security, such as creating home directories, setting password hash algorithms, and authorization. These settings are done independently of identity/user store settings.
For example, to create user home directories:
authconfig --enablemkhomedir --update
To set or change the hash algorithm used to encrypt user passwords:
authconfig --passalgo=sha512 --update

12.1.4.7. Configuring Fingerprint Authentication

There is one option to enable support for fingerprint readers. This option can be used alone or in conjunction with other authconfig settings, like LDAP user stores.
[root@server ~]# authconfig --enablefingerprint --update

12.1.4.8. Configuring Smart Card Authentication

All that is required to use smart cards with a system is to set the --enablesmartcard option:
[root@server ~]# authconfig --enablesmartcard --update
There are other configuration options for smart cards, such as changing the default smart card module, setting the behavior of the system when the smart card is removed, and requiring smart cards for login.
For example, this command instructs the system to lock out a user immediately if the smart card is removed (a setting of 1 ignores it if the smart card is removed):
[root@server ~]# authconfig --enablesmartcard --smartcardaction=0 --update
Once smart card authentication has been successfully configured and tested, then the system can be configured to require smart card authentication for users rather than simple password-based authentication.
[root@server ~]# authconfig --enablerequiresmartcard --update

Warning

Do not use the --enablerequiresmartcard option until you have successfully authenticated to the system using a smart card. Otherwise, users may be unable to log into the system.

12.1.4.9. Managing Kickstart and Configuration Files

The --update option updates all of the configuration files with the configuration changes. There are a couple of alternative options with slightly different behavior:
  • --kickstart writes the updated configuration to a kickstart file.
  • --test prints the full configuration, with changes, to stdout but does not edit any configuration files.
Additionally, authconfig can be used to back up and restore previous configurations. All archives are saved to a unique subdirectory in the /var/lib/authconfig/ directory. For example, the --savebackup option gives the backup directory as 2011-07-01:
[root@server ~]# authconfig --savebackup=2011-07-01
This backs up all of the authentication configuration files beneath the /var/lib/authconfig/backup-2011-07-01 directory.
Any of the saved backups can be used to restore the configuration using the --restorebackup option, giving the name of the manually-saved configuration:
[root@server ~]# authconfig --restorebackup=2011-07-01
Additionally, authconfig automatically makes a backup of the configuration before it applies any changes (with the --update option). The configuration can be restored from the most recent automatic backup, without having to specify the exact backup, using the --restorelastbackup option.

12.1.5. Using Custom Home Directories

If LDAP users have home directories that are not in /home and the system is configured to create home directories the first time users log in, then these directories are created with the wrong permissions.
  1. Apply the correct SELinux context and permissions from the /home directory to the home directory that is created on the local system. For example:
    [root@server ~]# semanage fcontext -a -e /home /home/locale
  2. Install the oddjob-mkhomedir package on the system.
    This package provides the pam_oddjob_mkhomedir.so library, which the Authentication Configuration Tool uses to create home directories. The pam_oddjob_mkhomedir.so library, unlike the default pam_mkhomedir.so library, can create SELinux labels.
    The Authentication Configuration Tool automatically uses the pam_oddjob_mkhomedir.so library if it is available. Otherwise, it will default to using pam_mkhomedir.so.
  3. Make sure the oddjobd service is running.
  4. Re-run the Authentication Configuration Tool and enable home directories, as in Section 12.1.3, “Configuring Alternative Authentication Features”.
If home directories were created before the home directory configuration was changed, then correct the permissions and SELinux contexts. For example:
[root@server ~]# semanage fcontext -a -e /home /home/locale
# restorecon -R -v /home/locale

12.2. Using and Caching Credentials with SSSD

The System Security Services Daemon (SSSD) provides access to different identity and authentication providers.

12.2.1. About SSSD

Most system authentication is configured locally, which means that services must check with a local user store to determine users and credentials. What SSSD does is allow a local service to check with a local cache in SSSD, but that cache may be taken from any variety of remote identity providers — an LDAP directory, an Identity Management domain, Active Directory, possibly even a Kerberos realm.
SSSD also caches those users and credentials, so if the local system or the identity provider go offline, the user credentials are still available to services to verify.
SSSD is an intermediary between local clients and any configured data store. This relationship brings a number of benefits for administrators:
  • Reducing the load on identification/authentication servers. Rather than having every client service attempt to contact the identification server directly, all of the local clients can contact SSSD which can connect to the identification server or check its cache.
  • Permitting offline authentication. SSSD can optionally keep a cache of user identities and credentials that it retrieves from remote services. This allows users to authenticate to resources successfully, even if the remote identification server is offline or the local machine is offline.
  • Using a single user account. Remote users frequently have two (or even more) user accounts, such as one for their local system and one for the organizational system. This is necessary to connect to a virtual private network (VPN). Because SSSD supports caching and offline authentication, remote users can connect to network resources simply by authenticating to their local machine and then SSSD maintains their network credentials.
Additional Resources
While this chapter covers the basics of configuring services and domains in SSSD, this is not a comprehensive resource. Many other configuration options are available for each functional area in SSSD; check out the man page for the specific functional area to get a complete list of options.
Some of the common man pages are listed in Table 12.1, “A Sampling of SSSD Man Pages”. There is also a complete list of SSSD man pages in the "See Also" section of the sssd(8) man page.

Table 12.1. A Sampling of SSSD Man Pages

Functional Area Man Page
General Configuration sssd.conf(8)
sudo Services sssd-sudo
LDAP Domains sssd-ldap
Active Directory Domains
sssd-ad
sssd-ldap
Identity Management (IdM or IPA) Domains
sssd-ipa
sssd-ldap
Kerberos Authentication for Domains sssd-krb5
OpenSSH Keys
sss_ssh_authorizedkeys
sss_ssh_knownhostsproxy
Cache Maintenance
sss_cache (cleanup)
sss_useradd, sss_usermod, sss_userdel, sss_seed (user cache entry management)

12.2.2. Setting up the sssd.conf File

SSSD services and domains are configured in a .conf file. By default, this is /etc/sssd/sssd.conf — although that file must be created and configured manually, since SSSD is not configured after installation.

12.2.2.1. Creating the sssd.conf File

There are three parts of the SSSD configuration file:
  • [sssd], for general SSSD process and operational configuration; this basically lists the configured services, domains, and configuration parameters for each
  • [service_name], for configuration options for each supported system service, as described in Section 12.2.4, “SSSD and System Services”
  • [domain_type/DOMAIN_NAME], for configuration options for each configured identity provider

    Important

    While services are optional, at least one identity provider domain must be configured before the SSSD service can be started.

Example 12.1. Simple sssd.conf File

[sssd]
domains = LOCAL
services = nss
config_file_version = 2

[nss]
filter_groups = root
filter_users = root

[domain/LOCAL]
id_provider = local
auth_provider = local
access_provider = permit

The [sssd] section has three important parameters:
  • domains lists all of the domains, configured in the sssd.conf, which SSSD uses as identity providers. If a domain is not listed in the domains key, it is not used by SSSD, even if it has a configuration section.
  • services lists all of the system services, configured in the sssd.conf, which use SSSD; when SSSD starts, the corresponding SSSD service is started for each configured system service. If a service is not listed in the services key, it is not used by SSSD, even if it has a configuration section.
  • config_file_version sets the version of the configuration file to set file format expectations. This is version 2, for all recent SSSD versions.

Note

Even if a service or domain is configured in the sssd.conf file, SSSD does not interact with that service or domain unless it is listed in the services or domains parameters, respectively, in the [sssd] section.
Other configuration parameters are listed in the sssd.conf man page.
Each service and domain parameter is described in its respective configuration section in this chapter and in their man pages.

12.2.2.2. Using a Custom Configuration File

By default, the sssd process assumes that the configuration file is /etc/sssd/sssd.conf.
An alternative file can be passed to SSSD by using the -c option with the sssd command:
[root@server ~]# sssd -c /etc/sssd/customfile.conf --daemon

12.2.3. Starting and Stopping SSSD

Important

Configure at least one domain before starting SSSD for the first time. See Section 12.2.10, “SSSD and Identity Providers (Domains)”.
Either the service command or the /etc/init.d/sssd script can start SSSD. For example:
[root@server ~]# service sssd start
By default, SSSD is not configured to start automatically. There are two ways to change this behavior:
  • Enabling SSSD through the authconfig command:
    [root@server ~]# authconfig --enablesssd --enablesssdauth --update
  • Adding the SSSD process to the start list using the chkconfig command:
    [root@server ~]# chkconfig sssd on

12.2.4. SSSD and System Services

SSSD and its associated services are configured in the sssd.conf file. The [sssd] section also lists the services that are active and should be started when sssd starts within the services directive.
SSSD can provide credentials caches for several system services:

12.2.5. Configuring Services: NSS

SSSD provides an NSS module, sssd_nss, which instructs the system to use SSSD to retrieve user information. The NSS configuration must include a reference to the SSSD module, and then the SSSD configuration sets how SSSD interacts with NSS.

12.2.5.1. About NSS Service Maps and SSSD

The Name Service Switch (NSS) provides a central configuration for services to look up a number of configuration and name resolution services. NSS provides one method of mapping system identities and services with configuration sources.
SSSD works with NSS as a provider services for several types of NSS maps:
  • Passwords (passwd)
  • User groups (shadow)
  • Groups (groups)
  • Netgroups (netgroups)
  • Services (services)

12.2.5.2. Configuring NSS Services to Use SSSD

NSS can use multiple identity and configuration providers for any and all of its service maps. The default is to use system files for services; for SSSD to be included, the nss_sss module has to be included for the desired service type.
  1. Use the Authentication Configuration tool to enable SSSD. This automatically configured the nsswitch.conf file to use SSSD as a provider.
    [root@server ~]# authconfig --enablesssd --update
    This automatically configures the password, shadow, group, and netgroups services maps to use the SSSD module:
    passwd:     files sss
    shadow:     files sss
    group:      files sss
    
    netgroup:   files sss
  2. The services map is not enabled by default when SSSD is enabled with authconfig. To include that map, open the nsswitch.conf file and add the sss module to the services map:
    [root@server ~]# vim /etc/nsswitch.conf
    
    ...
    services: file sss
    ...

12.2.5.3. Configuring SSSD to Work with NSS

The options and configuration that SSSD uses to service NSS requests are configured in the SSSD configuration file, in the [nss] services section.
  1. Open the sssd.conf file.
    [root@server ~]# vim /etc/sssd/sssd.conf
  2. Make sure that NSS is listed as one of the services that works with SSSD.
    [sssd]
    config_file_version = 2
    reconnection_retries = 3
    sbus_timeout = 30
    services = nss, pam
  3. In the [nss] section, change any of the NSS parameters. These are listed in Table 12.2, “SSSD [nss] Configuration Parameters”.
    [nss]
    filter_groups = root
    filter_users = root
    reconnection_retries = 3
    entry_cache_timeout = 300
    entry_cache_nowait_percentage = 75
  4. Restart SSSD.
    [root@server ~]# service sssd restart

Table 12.2. SSSD [nss] Configuration Parameters

Parameter Value Format Description
entry_cache_nowait_percentage integer Specifies how long sssd_nss should return cached entries before refreshing the cache. Setting this to zero (0) disables the entry cache refresh.
This configures the entry cache to update entries in the background automatically if they are requested if the time before the next update is a certain percentage of the next interval. For example, if the interval is 300 seconds and the cache percentage is 75, then the entry cache will begin refreshing when a request comes in at 225 seconds — 75% of the interval.
The allowed values for this option are 0 to 99, which sets the percentage based on the entry_cache_timeout value. The default value is 50%.
entry_negative_timeout integer Specifies how long, in seconds, sssd_nss should cache negative cache hits. A negative cache hit is a query for an invalid database entries, including non-existent entries.
filter_users, filter_groups string Tells SSSD to exclude certain users from being fetched from the NSS database. This is particularly useful for system accounts such as root.
filter_users_in_groups Boolean Sets whether users listed in the filter_users list appear in group memberships when performing group lookups. If set to FALSE, group lookups return all users that are members of that group. If not specified, this value defaults to true, which filters the group member lists.
debug_level integer, 0 - 9 Sets a debug logging level.

12.2.6. Configuring Services: PAM

Warning

A mistake in the PAM configuration file can lock users out of the system completely. Always back up the configuration files before performing any changes, and keep a session open so that any changes can be reverted.
SSSD provides a PAM module, sssd_pam, which instructs the system to use SSSD to retrieve user information. The PAM configuration must include a reference to the SSSD module, and then the SSSD configuration sets how SSSD interacts with PAM.
To configure the PAM service:
  1. Use authconfig to enable SSSD for system authentication.
    # authconfig --update --enablesssd --enablesssdauth
    This automatically updates the PAM configuration to reference all of the SSSD modules:
    #%PAM-1.0
    # This file is auto-generated.
    # User changes will be destroyed the next time authconfig is run.
    auth        required      pam_env.so
    auth        sufficient    pam_unix.so nullok try_first_pass
    auth        requisite     pam_succeed_if.so uid >= 500 quiet
    auth sufficient pam_sss.so use_first_pass
    auth        required      pam_deny.so
    
    account     required      pam_unix.so 
    account     sufficient    pam_localuser.so
    account     sufficient    pam_succeed_if.so uid < 500 quiet
    account [default=bad success=ok user_unknown=ignore] pam_sss.so
    account     required      pam_permit.so
    
    password    requisite     pam_cracklib.so try_first_pass retry=3
    password    sufficient    pam_unix.so sha512 shadow nullok try_first_pass use_authtok
    password sufficient pam_sss.so use_authtok
    password    required      pam_deny.so
    
    session     optional      pam_keyinit.so revoke
    session     required      pam_limits.so
    session     [success=1 default=ignore] pam_succeed_if.so service in crond quiet use_uid
    session sufficient pam_sss.so
    session     required      pam_unix.so
    
    These modules can be set to include statements, as necessary.
  2. Open the sssd.conf file.
    # vim /etc/sssd/sssd.conf
  3. Make sure that PAM is listed as one of the services that works with SSSD.
    [sssd]
    config_file_version = 2
    reconnection_retries = 3
    sbus_timeout = 30
    services = nss, pam
  4. In the [pam] section, change any of the PAM parameters. These are listed in Table 12.3, “SSSD [pam] Configuration Parameters”.
    [pam]
    reconnection_retries = 3
    offline_credentials_expiration = 2
    offline_failed_login_attempts = 3
    offline_failed_login_delay = 5
  5. Restart SSSD.
    [root@server ~]# service sssd restart

Table 12.3. SSSD [pam] Configuration Parameters

Parameter Value Format Description
offline_credentials_expiration integer Sets how long, in days, to allow cached logins if the authentication provider is offline. This value is measured from the last successful online login. If not specified, this defaults to zero (0), which is unlimited.
offline_failed_login_attempts integer Sets how many failed login attempts are allowed if the authentication provider is offline. If not specified, this defaults to zero (0), which is unlimited.
offline_failed_login_delay integer Sets how long to prevent login attempts if a user hits the failed login attempt limit. If set to zero (0), the user cannot authenticate while the provider is offline once he hits the failed attempt limit. Only a successful online authentication can re-enable offline authentication. If not specified, this defaults to five (5).

12.2.7. Configuring Services: autofs

12.2.7.1. About Automount, LDAP, and SSSD

Automount maps are commonly flat files, which define a relationship between a map, a mount directory, and a fileserver. (Automount is described in the Storage Administration Guide.)
For example, let's say that there is a fileserver called nfs.example.com which hosts the directory pub, and automount is configured to mount directories in the /shares/ directory. So, the mount location is /shares/pub. All of the mounts are listed in the auto.master file, which identifies the different mount directories and the files which configure them. The auto.shares file then identifies each file server and mount directory which goes into the /shares/ directory. The relationships could be viewed like this:
        auto.master
   _________|__________
   |                 |
   |                 |
/shares/        auto.shares
                     |
		     |
		     |
            nfs.example.com:pub
Every mount point, then, is defined in two different files (at a minimum): the auto.master and auto.whatever file, and those files have to be available to each local automount process.
One way for administrators to manage that for large environments is to store the automount configuration in a central LDAP directory, and just configure each local system to point to that LDAP directory. That means that updates only need to be made in a single location, and any new maps are automatically recognized by local systems.
For automount-LDAP configuration, the automount files are stored as LDAP entries, which are then translated into the requisite automount files. Each element is then translated into an LDAP attribute.
The LDAP entries look like this:
# container entry
dn: cn=automount,dc=example,dc=com
objectClass: nsContainer
objectClass: top
cn: automount

# master map entry
dn: automountMapName=auto.master,cn=automount,dc=example,dc=com
objectClass: automountMap
objectClass: top
automountMapName: auto.master

# shares map entry
dn: automountMapName=auto.shares,cn=automount,dc=example,dc=com
objectClass: automountMap
objectClass: top
automountMapName: auto.shares

# shares mount point
dn: automountKey=/shares,automountMapName=auto.master,cn=automount,dc=example,dc=com
objectClass: automount
objectClass: top
automountKey: /shares
automountInformation: auto.shares

# pub mount point
dn: automountKey=pub,automountMapName=auto.shares,cn=automount,dc=example,dc=com
objectClass: automount
objectClass: top
automountKey: pub
automountInformation: filer.example.com:/pub
description: pub
The schema elements, then, match up to the structure like this (with the RFC 2307 schema):
	              auto.master
                      objectclass: automountMap
                      filename attribute: automountMapName
   _______________________|_________________________
   |                                               |
   |                                               |
/shares/                                       auto.shares
objectclass: automount                         objectclass: automountMap
mount point name attribute: automountKey       filename attribute: automountMapName
map name attribute: automountInformation           |
		                                   |
                                                   |
                                          nfs.example.com:pub
                                          objectclass: automount
                                          mount point name attribute: automountKey
                                          fileserver attribute: automountInformation
autofs uses those schema elements to derive the automount configuration. The /etc/sysconfig/autofs file identifies the LDAP server, directory location, and schema elements used for automount entities:
LDAP_URI=ldap://ldap.example.com
SEARCH_BASE="cn=automount,dc=example,dc=com"
MAP_OBJECT_CLASS="automountMap"
ENTRY_OBJECT_CLASS="automount"
MAP_ATTRIBUTE="automountMapName"
ENTRY_ATTRIBUTE="automountKey"
VALUE_ATTRIBUTE="automountInformation"
Rather than pointing the automount configuration to the LDAP directory, it can be configured to point to SSSD. SSSD, then, stores all of the information that automount needs, and as a user attempts to mount a directory, that information is cached into SSSD. This offers several advantages for configuration — such as failover, service discovery, and timeouts — as well as performance improvements by reducing the number of connections to the LDAP server. Most important, using SSSD allows all mount information to be cached, so that clients can still successfully mount directories even if the LDAP server goes offline.

12.2.7.2. Configuring autofs Services in SSSD

  1. Make sure that the autofs and libsss_autofs packages are installed.
  2. Open the sssd.conf file.
    [root@server ~]# vim /etc/sssd/sssd.conf
  3. Add the autofs service to the list of services that SSSD manages.
    [sssd]
    services = nss,pam,autofs
    ....
  4. Create a new [autofs] service configuration section. This section can be left blank; there is only one configurable option, for timeouts for negative cache hits.
    This section is required, however, for SSSD to recognize the autofs service and supply the default configuration.
    [autofs]
    
    
    
  5. The automount information is read from a configured LDAP domain in the SSSD configuration, so an LDAP domain must be available. If no additional settings are made, then the configuration defaults to the RFC 2307 schema and the LDAP search base (ldap_search_base) for the automount information. This can be customized:
    • The directory type, autofs_provider; this defaults to the id_provider value; a value of none explicitly disables autofs for the domain.
    • The search base, ldap_autofs_search_base.
    • The object class to use to recognize map entries, ldap_autofs_map_object_class
    • The attribute to use to recognize map names, ldap_autofs_map_name
    • The object class to use to recognize mount point entries, ldap_autofs_entry_object_class
    • The attribute to use to recognize mount point names, ldap_autofs_entry_key
    • The attribute to use for additional configuration information for the mount point, ldap_autofs_entry_value
    For example:
    [domain/LDAP]
    ...
    autofs_provider=ldap
    ldap_autofs_search_base=cn=automount,dc=example,dc=com
    ldap_autofs_map_object_class=automountMap
    ldap_autofs_entry_object_class=automount
    ldap_autofs_map_name=automountMapName
    ldap_autofs_entry_key=automountKey
    ldap_autofs_entry_value=automountInformation
  6. Save and close the sssd.conf file.
  7. Configure autofs to look for the automount map information in SSSD by editing the nsswitch.conf file and changing the location from ldap to sss:
    [root@server ~]# vim /etc/nsswitch.conf
    
    automount: files sss
  8. Restart SSSD.
    [root@server ~]# service sssd restart

12.2.8. Configuring Services: sudo

12.2.8.1. About sudo, LDAP, and SSSD

sudo rules are defined in the sudoers file, which must be distributed separately to every machine to maintain consistency.
One way for administrators to manage that for large environments is to store the sudo configuration in a central LDAP directory, and just configure each local system to point to that LDAP directory. That means that updates only need to be made in a single location, and any new rules are automatically recognized by local systems.
For sudo-LDAP configuration, each sudo rule is stored as an LDAP entry, with each component of the sudo rule defined in an LDAP attribute.
The sudoers rule looks like this:
Defaults    env_keep+=SSH_AUTH_SOCK
...
%wheel        ALL=(ALL)       ALL
The LDAP entry looks like this:
# sudo defaults
dn: cn=defaults,ou=SUDOers,dc=example,dc=com
objectClass: top
objectClass: sudoRole
cn: defaults
description: Default sudoOptions go here
sudoOption: env_keep+=SSH_AUTH_SOCK

# sudo rule
dn: cn=%wheel,ou=SUDOers,dc=example,dc=com
objectClass: top
objectClass: sudoRole
cn: %wheel
sudoUser: %wheel
sudoHost: ALL
sudoCommand: ALL

Note

SSSD only caches sudo rules which apply to the local system, depending on the value of the sudoHost attribute. This can mean that the sudoHost value is set to ALL, uses a regular expression that matches the hostname, matches the systems netgroup, or matches the systems hostname, fully-qualified domain name, or IP address.
The sudo service can be configured to point to an LDAP server and to pull its rule configuration from those LDAP entries. Rather than pointing the sudo configuration to the LDAP directory, it can be configured to point to SSSD. SSSD, then, stores all of the information that sudo needs, and every time a user attempts a sudo-related operation, the latest sudo configuration can be pulled from the LDAP directory (through SSSD). SSSD, however, also caches all of the sudo riles, so that users can perform tasks, using that centralized LDAP configuration, even if the LDAP server goes offline.

12.2.8.2. Configuring sudo with SSSD

All of the SSSD sudo configuration options are listed in the sssd-ldap(5) man page.
To configure the sudo service:
  1. Open the sssd.conf file.
    [root@server ~]# vim /etc/sssd/sssd.conf
  2. Add the sudo service to the list of services that SSSD manages.
    [sssd]
    services = nss,pam,sudo
    ....
  3. Create a new [sudo] service configuration section. This section can be left blank; there is only one configurable option, for evaluating the sudo not before/after period.
    This section is required, however, for SSSD to recognize the sudo service and supply the default configuration.
    [sudo]
    
    
    
  4. The sudo information is read from a configured LDAP domain in the SSSD configuration, so an LDAP domain must be available. For an LDAP provider, these parameters are required:
    • The directory type, sudo_provider; this is always ldap.
    • The search base, ldap_sudo_search_base.
    • The URI for the LDAP server, ldap_uri.
    For example:
    [domain/LDAP]
    id_provider = ldap
    
    sudo_provider = ldap
    ldap_uri = ldap://example.com
    ldap_sudo_search_base = ou=sudoers,dc=example,dc=com
    For an Identity Management (IdM or IPA) provider, there are additional parameters required to perform Kerberos authentication when connecting to the server.
    [domain/IDM]
    id_provider = ipa
    ipa_domain = example.com
    ipa_server = ipa.example.com
    ldap_tls_cacert = /etc/ipa/ca.crt
    
    sudo_provider = ldap
    ldap_uri = ldap://ipa.example.com
    ldap_sudo_search_base = ou=sudoers,dc=example,dc=com
    ldap_sasl_mech = GSSAPI
    ldap_sasl_authid = host/hostname.example.com
    ldap_sasl_realm = EXAMPLE.COM
    krb5_server = ipa.example.com

    Note

    The sudo_provider type for an Identity Management provider is still ldap.
  5. Set the intervals to use to refresh the sudo rule cache.
    The cache for a specific system user is always checked and updated whenever that user performs a task. However, SSSD caches all rules which relate to the local system. That complete cache is updated in two ways:
    • Incrementally, meaning only changes to rules since the last full update (ldap_sudo_smart_refresh_interval, the time in seconds); the default is 15 minutes,
    • Fully, which dumps the entire caches and pulls in all of the current rules on the LDAP server(ldap_sudo_full_refresh_interval, the time in seconds); the default is six hours.
    These two refresh intervals are set separately. For example:
    [domain/LDAP]
    ...
    ldap_sudo_full_refresh_interval=86400
    ldap_sudo_smart_refresh_interval=3600

    Note

    SSSD only caches sudo rules which apply to the local system. This can mean that the sudoHost value is set to ALL, uses a regular expression that matches the hostname, matches the systems netgroup, or matches the systems hostname, fully-qualified domain name, or IP address.
  6. Optionally, set any values to change the schema used for sudo rules.
    Schema elements are set in the ldap_sudorule_* attributes. By default, all of the schema elements use the schema defined in sudoers.ldap; these defaults will be used in almost all deployments.
  7. Save and close the sssd.conf file.
  8. Configure sudo to look for rules configuration in SSSD by editing the nsswitch.conf file and adding the sss location:
    [root@server ~]# vim /etc/nsswitch.conf
    
    sudoers: files sss
  9. Restart SSSD.
    [root@server ~]# service sssd restart

12.2.9. Configuring Services: OpenSSH and Cached Keys

OpenSSH creates secure, encrypted connections between two systems. One machine authenticates to another machine to allow access; the authentication can be of the machine itself for server connections or of a user on that machine. OpenSSH is described in more detail in Chapter 13, OpenSSH.
This authentication is performed through public-private key pairs that identify the authenticating user or machine. The remote machine or user attempting to access the machine presents a key pair. The local machine then elects whether to trust that remote entity; if it is trusted, the public key for that remote machine is stored in the known_hosts file or for the remote user in authorized_keys. Whenever that remote machine or user attempts to authenticate again, the local system simply checks the known_hosts or authorized_keys file first to see if that remote entity is recognized and trusted. If it is, then access is granted.
The first problem comes in verifying those identities reliably.
The known_hosts file is a triplet of the machine name, its IP address, and its public key:
server.example.com,255.255.255.255 ssh-rsa AbcdEfg1234ZYX098776/AbcdEfg1234ZYX098776/AbcdEfg1234ZYX098776=
The known_hosts file can quickly become outdated for a number of different reasons: systems using DHCP cycle through IP addresses, new keys can be re-issued periodically, or virtual machines or services can be brought online and removed. This changes the hostname, IP address, and key triplet.
Administrators have to clean and maintain a current known_hosts file to maintain security. (Or system users get in the habit of simply accepting any machine and key presented, which negates the security benefits of key-based security.)
Additionally, a problem for both machines and users is distributing keys in a scalable way. Machines can send their keys as part of establishing an encrypted session, but users have to supply their keys in advance. Simply propagating and then updating keys consistently is a difficult administrative task.
Lastly, SSH key and machine information are only maintained locally. There may be machines or users on the network which are recognized and trusted by some systems and not by others because the known_hosts file has not been updated uniformly.
The goal of SSSD is to server as a credentials cache. This includes working as a credentials cache for SSH public keys for machines and users. OpenSSH is configured to reference SSSD to check for cached keys; SSSD uses Red Hat Linux's Identity Management (IPA) domain as an identity, and Identity Management actually stores the public keys and host information.

NOTE

Only Linux machines enrolled, or joined, in the Identity Management domain can use SSSD as a key cache for OpenSSH. Other Unix machines and Windows machines must use the regular authentication mechanisms with the known_hosts file.

12.2.9.1. Configuring OpenSSH to Use SSSD for Host Keys

OpenSSH is configured in either a user-specific configuration file (~/.ssh/config) or a system-wide configuration file (/etc/ssh/ssh_config). The user file has precedence over the system settings and the first obtained value for a parameter is used. The formatting and conventions for this file are covered in Chapter 13, OpenSSH.
In order to manage host keys, SSSD has a tool, sss_ssh_knownhostsproxy, which performs three operations:
  1. Retrieves the public host key from the enrolled Linux system.
  2. Stores the host key in a custom hosts file, .ssh/sss_known_hosts.
  3. Establishes a connection with the host machine, either a socket (the default) or a secure connection.
This tool has the format:
sss_ssh_knownhostsproxy [-d sssd_domain] [-p ssh_port] HOST [PROXY_COMMAND]

Table 12.4. sss_ssh_knownhostsproxy Options

Short Argument Long Argument Description
HOSTNAME Gives the hostname of the host to check and connect to. In the OpenSSH configuration file, this can be a token, %h.
PROXY_COMMAND Passes a proxy command to use to connect to the SSH client. This is similar to running ssh -o ProxyCommand=value. This option is used when running sss_ssh_knownhostsproxy from the command line or through another script, but is not necessary in the OpenSSH configuration file.
-d sssd_domain --domain sssd_domain Only searches for public keys in entries in the specified domain. If not given, SSSD searches for keys in all configured domains.
-p port --port port Uses this port to connect to the SSH client. By default, this is port 22.

To use this SSSD tool, add or edit two parameters to the ssh_config or ~/.ssh/config file:
  • Specify the command to use to connect to the SSH client (ProxyCommand). This is the sss_ssh_knownhostsproxy, with the desired arguments and hostname.
  • Specify the location of the SSSD hosts file, rather than the default known_hosts file (UserKnownHostsFile). The SSSD hosts file is .ssh/sss_known_hosts.
For example, this looks for public keys in the IPA1 SSSD domain and connects over whatever port and host are supplied:
ProxyCommand /usr/bin/sss_ssh_knownhostsproxy -p %p -d IPA1 %h
UserKnownHostsFile2 .ssh/sss_known_hosts

12.2.9.2. Configuring OpenSSH to Use SSSD for User Keys

User keys are stored on a local system in the authorized_keys file for OpenSSH. As with hosts, SSSD can maintain and automatically update a separate cache of user public keys for OpenSSH to refer to. This is kept in the .ssh/sss_authorized_keys file.
OpenSSH is configured in either a user-specific configuration file (~/.ssh/config) or a system-wide configuration file (/etc/ssh/ssh_config). The user file has precedence over the system settings and the first obtained value for a parameter is used. The formatting and conventions for this file are covered in Chapter 13, OpenSSH.
In order to manage user keys, SSSD has a tool, sss_ssh_authorizedkeys, which performs two operations:
  1. Retrieves the user's public key from the user entries in the Identity Management (IPA) domain.
  2. Stores the user key in a custom file, .ssh/sss_authorized_keys, in the standard authorized keys format.
This tool has the format:
sss_ssh_authorizedkeys [-d sssd_domain] USER

Table 12.5. sss_ssh_authorizedkeys Options

Short Argument Long Argument Description
USER Gives the username or account name for which to obtain the public key. In the OpenSSH configuration file, this can be represented by a token, %u.
-d sssd_domain --domain sssd_domain Only searches for public keys in entries in the specified domain. If not given, SSSD searches for keys in all configured domains.

There are two possible options for how to configure OpenSSH to use SSSD for user keys, depending on the SSH deployment:
  • Most commonly, SSH supports the authorized key command. In that case, it is necessary only to specify the command to run to retrieve user keys. For example:
    AuthorizedKeysCommand /usr/bin/sss_ssh_authorizedkeys
  • SSH can also support a public key agent. In that case, give the command to use to retrieve agent keys, including tokens for required arguments (such as the username):
    PubKeyAgent /usr/bin/sss_ssh_authorizedkeys %u

12.2.10. SSSD and Identity Providers (Domains)

SSSD recognizes domains, which are entries within the SSSD configuration file associated with different, external data sources. Domains are a combination of an identity provider (for user information) and, optionally, other providers such as authentication (for authentication requests) and for other operations, such as password changes. (The identity provider can also be used for all operations, if all operations are performed within a single domain or server.)
SSSD works with different LDAP identity providers (including OpenLDAP, Red Hat Directory Server, and Microsoft Active Directory) and can use native LDAP authentication, Kerberos authentication, or provider-specific authentication protocols (such as Active Directory).
A domain configuration defines the identity provider, the authentication provider, and any specific configuration to access the information in those providers. There are several types of identity and authentication providers:
  • LDAP, for general LDAP servers
  • Active Directory (an extension of the LDAP provider type)
  • Identity Management (an extension of the LDAP provider type)
  • Local, for the local SSSD database
  • Proxy
  • Kerberos (authentication provider only)
The identity and authentication providers can be configured in different combinations in the domain entry. The possible combinations are listed in Table 12.6, “Identity Store and Authentication Type Combinations”.

Table 12.6. Identity Store and Authentication Type Combinations

Identification Provider Authentication Provider
Identity Management (LDAP) Identity Management (LDAP)
Active Directory (LDAP) Active Directory (LDAP)
Active Directory (LDAP) Kerberos
LDAP LDAP
LDAP Kerberos
proxy LDAP
proxy Kerberos
proxy proxy

Along with the domain entry itself, the domain name must be added to the list of domains that SSSD will query. For example:
[sssd]
domains = LOCAL,Name
...

[domain/Name]
id_provider = type
auth_provider = type
provider_specific = value
global = value
global attributes are available to any type of domain, such as cache and time out settings. Each identity and authentication provider has its own set of required and optional configuration parameters.

Table 12.7. General [domain] Configuration Parameters

Parameter Value Format Description
id_provider string Specifies the data backend to use for this domain. The supported identity backends are:
  • ldap
  • ipa (Identity Management in Red Hat Enterprise Linux)
  • ad (Microsoft Active Directory)
  • proxy, for a legacy NSS provider, such as nss_nis. Using a proxy ID provider also requires specifying the legacy NSS library to load to start successfully, set in the proxy_lib_name option.
  • local, the SSSD internal local provider
auth_provider string Sets the authentication provider used for the domain. The default value for this option is the value of id_provider. The supported authentication providers are ldap, ipa, ad, krb5 (Kerberos), proxy, and none.
min_id,max_id integer Optional. Specifies the UID and GID range for the domain. If a domain contains entries that are outside that range, they are ignored. The default value for min_id is 1; the default value for max_id is 0, which is unlimited.

Important

The default min_id value is the same for all types of identity provider. If LDAP directories are using UID numbers that start at one, it could cause conflicts with users in the local /etc/passwd file. To avoid these conflicts, set min_id to 1000 or higher as possible.
cache_credentials Boolean Optional. Specifies whether to store user credentials in the local SSSD domain database cache. The default value for this parameter is false. Set this value to true for domains other than the LOCAL domain to enable offline authentication.
entry_cache_timeout integer Optional. Specifies how long, in seconds, SSSD should cache positive cache hits. A positive cache hit is a successful query.
use_fully_qualified_names Boolean Optional. Specifies whether requests to this domain require fully-qualified domain names. If set to true, all requests to this domain must use fully-qualified domain names. It also means that the output from the request displays the fully-qualified name. Restricting requests to fully-qualified user names allows SSSD to differentiate between domains with users with conflicting usernames.
If use_fully_qualified_names is set to false, it is possible to use the fully-qualified name in the requests, but only the simplified version is displayed in the output.
SSSD can only parse names based on the domain name, not the realm name. The same name can be used for both domains and realms, however.

12.2.11. Creating Domains: LDAP

An LDAP domain simply means that SSSD uses an LDAP directory as the identity provider (and, optionally, also as an authentication provider). SSSD supports several major directory services:
  • Red Hat Directory Server
  • OpenLDAP
  • Identity Management (IdM or IPA)
  • Microsoft Active Directory 2008 R2

Note

All of the parameters available to a general LDAP identity provider are also available to Identity Management and Active Directory identity providers, which are subsets of the LDAP provider.

12.2.11.1. Parameters for Configuring an LDAP Domain

An LDAP directory can function as both an identity provider and an authentication provider. The configuration requires enough information to identify and connect to the user directory in the LDAP server, but the way that those connection parameters are defined is flexible.
Other options are available to provide more fine-grained control, like specifying a user account to use to connect to the LDAP server or using different LDAP servers for password operations. The most common options are listed in Table 12.8, “LDAP Domain Configuration Parameters”.

Tip

Many other options are listed in the man page for LDAP domain configuration, sssd-ldap(5).

Table 12.8. LDAP Domain Configuration Parameters

Parameter Description
ldap_uri Gives a comma-separated list of the URIs of the LDAP servers to which SSSD will connect. The list is given in order of preference, so the first server in the list is tried first. Listing additional servers provides failover protection. This can be detected from the DNS SRV records if it is not given.
ldap_search_base Gives the base DN to use for performing LDAP user operations.
ldap_tls_reqcert Specifies how to check for SSL server certificates in a TLS session. There are four options:
  • never disables requests for certificates.
  • allow requests a certificate, but proceeds normally even if no certificate is given or a bad certificate is given.
  • try requests a certificate and proceeds normally if no certificate is given, If a bad certificate is given, the session terminates.
  • demand and hard are the same option. This requires a valid certificate or the session is terminated.
The default is hard.
ldap_tls_cacert Gives the full path and file name to the file that contains the CA certificates for all of the CAs that SSSD recognizes. SSSD will accept any certificate issued by these CAs.
This uses the OpenLDAP system defaults if it is not given explicitly.
ldap_referrals Sets whether SSSD will use LDAP referrals, meaning forwarding queries from one LDAP database to another. SSSD supports database-level and subtree referrals. For referrals within the same LDAP server, SSSD will adjust the DN of the entry being queried. For referrals that go to different LDAP servers, SSSD does an exact match on the DN. Setting this value to true enables referrals; this is the default.
Referrals can negatively impact overall performance because of the time spent attempting to trace referrals. Disabling referral checking can significantly improve performance.
ldap_schema Sets what version of schema to use when searching for user entries. This can be rfc2307, rfc2307bis, ad, or ipa. The default is rfc2307.
In RFC 2307, group objects use a multi-valued attribute, memberuid, which lists the names of the users that belong to that group. In RFC 2307bis, group objects use the member attribute, which contains the full distinguished name (DN) of a user or group entry. RFC 2307bis allows nested groups using the member attribute. Because these different schema use different definitions for group membership, using the wrong LDAP schema with SSSD can affect both viewing and managing network resources, even if the appropriate permissions are in place.
For example, with RFC 2307bis, all groups are returned when using nested groups or primary/secondary groups.
$ id
uid=500(myserver) gid=500(myserver) groups=500(myserver),510(myothergroup)
If SSSD is using RFC 2307 schema, only the primary group is returned.
This setting only affects how SSSD determines the group members. It does not change the actual user data.
ldap_search_timeout Sets the time, in seconds, that LDAP searches are allowed to run before they are canceled and cached results are returned.
When an LDAP search times out, SSSD automatically switches to offline mode.
ldap_network_timeout Sets the time, in seconds, SSSD attempts to poll an LDAP server after a connection attempt fails. The default is six seconds.
ldap_opt_timeout Sets the time, in seconds, to wait before aborting synchronous LDAP operations if no response is received from the server. This option also controls the timeout when communicating with the KDC in case of a SASL bind. The default is five seconds.

12.2.11.2. LDAP Domain Example

The LDAP configuration is very flexible, depending on your specific environment and the SSSD behavior. These are some common examples of an LDAP domain, but the SSSD configuration is not limited to these examples.

Note

Along with creating the domain entry, add the new domain to the list of domains for SSSD to query in the sssd.conf file. For example:
domains = LOCAL,LDAP1,AD,PROXYNIS

Example 12.2. A Basic LDAP Domain Configuration

An LDAP domain requires three things:
  • An LDAP server
  • The search base
  • A way to establish a secure connection
The last item depends on the LDAP environment. SSSD requires a secure connection since it handles sensitive information. This connection can be a dedicated TLS/SSL connection or it can use Start TLS.
Using a dedicated TLS/SSL connection simply uses an LDAPS connection to connect to the server and is therefore set as part of the ldap_uri option:
# An LDAP domain
[domain/LDAP]
cache_credentials = true

id_provider = ldap
auth_provider = ldap

ldap_uri = ldaps://ldap.example.com:636
ldap_search_base = dc=example,dc=com
Using Start TLS requires a way to input the certificate information to establish a secure connection dynamically over an insecure port. This is done using the ldap_id_use_start_tls option to use Start TLS and then ldap_tls_cacert to identify the CA certificate which issued the SSL server certificates.
# An LDAP domain
[domain/LDAP]
cache_credentials = true

id_provider = ldap
auth_provider = ldap

ldap_uri = ldap://ldap.example.com
ldap_search_base = dc=example,dc=com
ldap_id_use_start_tls = true
ldap_tls_reqcert = demand
ldap_tls_cacert = /etc/pki/tls/certs/ca-bundle.crt

12.2.12. Creating Domains: Identity Management (IdM)

The Identity Management (IdM or IPA) identity provider is an extension of a generic LDAP provider. All of the configuration options for an LDAP provider are available to the IdM provider, as well as some additional parameters which allow SSSD to work as a client of the IdM domain and extend IdM functionality.
Identity Management can work as a provider for identities, authentication, access control rules, and passwords, all of the *_provider parameters for a domain. Additionally, Identity Management has configuration options within its own domain to manage SELinux policies, automount information, and host-based access control. All of those features in IdM domains can be tied to SSSD configuraiton, allowing those security-related policies to be applied and cached for system users.

Example 12.3. Basic IdM Provider

An IdM provider, like an LDAP provider, can be set to serve several different services, including identity, authentication, and access control
For IdM servers, there are two additional settings which are very useful (although not required):
  • Use the specific IdM schema rather than the default RFC 2307 schema.
  • Set SSSD to update the Identity Management domain's DNS server with the IP address of this client when the client first connects to the IdM domain.
[sssd]
domains = local,example.com
...

[domain/example.com]
id_provider = ipa
ipa_server = ipaserver.example.com
ipa_hostname = ipa1.example.com
auth_provider = ipa
access_provider = ipa
chpass_provider = ipa

# set which schema to use
ldap_schema = ipa

# automatically update IdM DNS records
ipa_dyndns_update = true

Identity Management defines and maintains security policies and identities for users across a Linux domain. This includes access control policies, SELinux policies, and other rules. Some of these elements in the IdM domain interact directly with SSSD, using SSSD as an IdM client — and those features can be managed in the IdM domain entry in sssd.conf.
Most of the configuration parameters relate to setting schema elements (which is not relevant in most deployments because IdM uses a fixed schema) and never need to be changed. In fact, none of the features in IdM require client-side settings. But there may be circumstances where tweaking the behavior is helpful.

Example 12.4. IdM Provider with SELinux

IdM can define SELinux user policies for system users, so it can work as an SELinux provider for SSSD. This is set in the selinux_provider parameter. The provider defaults to the id_provider value, so this is not necessary to set explicitly to support SELinux rules. However, it can be useful to explicitly disable SELinux support for the IdM provider in SSSD.
selinux_provider = ipa

Example 12.5. IdM Provider with Host-Base Access Control

IdM can define host-based access controls, restricting access to services or entire systems based on what host a user is using to connect or attempting to connect to. This rules can be evaluated and enforced by SSSD as part of the access provider behavior.
For host-based access controls to be in effect, the Identity Management server must be the access provider, at a minimum.
There are two options which can be set for how SSSD evaluates host-based access control rules:
  • SSSD can evaluate what machine (source host) the user is using to connect to the IdM resource; this is disabled by default, so that only the target host part of the rule is evaluated.
  • SSSD can refresh the host-based access control rules in its cache at a specified interval.
For example:
access_provider = ipa	
ipa_hbac_refresh = 120

# check for source machine rules; disabled by default
ipa_hbac_support_srchost = true

Example 12.6. Identity Management with Cross-Realm Kerberos Trusts

Identity Management (IdM or IPA) can be configured with trusted relationships between Active Directory DNS domains and Kerberos realms. This allows Active Directory users to access services and hosts on Linux systems.
There are two configuration settings in SSSD that are used with cross-realm trusts:
  • A service that adds required data to Kerberos tickets
  • A setting to support subdomains
Kerberos Ticket Data
Microsoft uses a special authorization structure called privileged access certificates or MS-PAC. A PAC is embedded in a Kerberos ticket as a way of identifying the entity to other Windows clients and servers in the Windows domain.
SSSD has a special PAC service which generates the additional data for Kerberos tickets. When using an Active Directory domain, it may be necessary to include the PAC data for Windows users. In that case, enable the pac service in SSSD:
[sssd]
services = nss, pam, pac
...
Windows Subdomains
Normally, a domain entry in SSSD corresponds directly to a single identity provider. However, with IdM cross-realm trusts, the IdM domain can trust another domain, so that the domains are transparent to each other. SSSD can follow that trusted relationship, so that if an IdM domain is configured, any Windows domain is also automatically searched and supported by SSSD — without having to be configured in a domain section in SSSD.
This is configured by adding the subdomains_provider parameter to the IdM domain section. This is actually an optional parameter; if a subdomain is discovered, then SSSD defaults to using the ipa provider type. However, this parameter can also be used to disable subdomain fetches by setting a value of none.
[domain/IDM]
...	
subdomains_provider = ipa
get_domains_timeout = 300

12.2.13. Creating Domains: Active Directory

The Active Directory identity provider is an extension of a generic LDAP provider. All of the configuration options for an LDAP provider are available to the Active Directory provider, as well as some additional parameters related to user accounts and identity mapping between Active Directory and system users.
There are some fundamental differences between standard LDAP servers and an Active Directory server. When configuring an Active Directory provider, there are some configuration areas, then, which require specific configuration:
  • Identities using a Windows security ID must be mapped to the corresponding Linux system user ID.
  • Searches must account for the range retrieval extension.
  • There may be performance issues with LDAP referrals.

12.2.13.1. Mapping Active Directory Securiy IDs and Linux User IDs

There are inherent structural differences between how Windows and Linux handle system users and in the user schemas used in Active Directory and standard LDAPv3 directory services. When using an Active Directory identity provider with SSSD to manage system users, it is necessary to reconcile the Active Directory-style user to the new SSSD user. There are two ways to do this:
  • Using Services for Unix to insert POSIX attributes on Windows user and group entries, and then having those attributes pulled into PAM/NSS
  • Using ID mapping on SSSD to create a map between Active Directory security IDs (SIDs) and the generated UIDs on Linux
ID mapping is the simplest option for most environments because it requires no additional packages or configuration on Active Directory.
12.2.13.1.1. The Mechanism of ID Mapping
Linux/Unix systems use a local user ID number and group ID number to identify users on the system. These UID:GID numbers are a simple integer, such as 501:501. These numbers are simple because they are always created and administered locally, even for systems which are part of a larger Linux/Unix domain.
Microsoft Windows and Active Directory use a different user ID structure to identify users, groups, and machines. Each ID is constructed of different segments that identify the security version, the issuing authority type, the machine, and the identity itself. For example:
S-1-5-21-3623811015-3361044348-30300820-1013
The third through sixth blocks are the machine identifier:
S-1-5-21-3623811015-3361044348-30300820-1013
The last block is the relative identifier (RID) which identifies the specific entity:
S-1-5-21-3623811015-3361044348-30300820-1013
A range of possible ID numbers are always assigned to SSSD. (This is a local range, so it is the same for every machine.)
|_____________________________|
|                             |
minimum ID                    max ID
This range is divided into 10,000 sections (by default), with each section allocated 200,000 IDs.
| slice 1 | slice 2 |   ...   |
|_________|_________|_________|
|         |         |         |
minimum ID                    max ID
When a new Active Directory domain is detected, the SID is hashed. Then, SSSD takes the modulus of the hash and the number of available sections to determine which ID section to assign to the Active Directory domain. This is a reliably consistent means of assigning ID sections, so the same ID range is assigned to the same Active Directory domain on most client machines.
| Active  | Active  |         |
|Directory|Directory|         |
|domain 1 |domain 2 |   ...   |
|         |         |         |
| slice 1 | slice 2 |   ...   |
|_________|_________|_________|
|         |         |         |
minimum ID                    max ID

Note

While the method of assigning ID sections is consistent, ID mapping is based on the order that an Active Directory domain is encountered on a client machine — so it may not result in consistent ID range assignments on all Linux client machines. If consistency is required, then consider disabling ID mapping and using explicit POSIX attributes.
12.2.13.1.2. ID Mapping Parameters
ID mapping is enabled in two parameters, one to enable the mapping and one to load the appropriate Active Directory user schema:
ldap_id_mapping = True
ldap_schema = ad

Note

When ID mapping is enabled, the uidNumber and gidNumber attributes are ignored. This prevents any manually-assigned values. If any values must be manually assigned, then all values must be manually assigned, and ID mapping should be disabled.
12.2.13.1.3. Mapping Users
When an Active Directory user attempts to log into a local system service for the first time, an entry for that user is created in the SSSD cache. The remote user is set up much like a system user:
  • A system UID is created for the user based on his SID and the ID range for that domain.
  • A GID is created for the user, which is identical to the UID.
  • A private group is created for the user.
  • A home directory is created, based on the home directory format in the sssd.conf file.
  • A shell is created, according to the system defaults or the setting in the sssd.conf file.
  • If the user belongs to any groups in the Active Directory domain, then, using the SID, SSSD adds the user to those groups on the Linux system.

12.2.13.2. Active Directory Users and Range Retrieval Searches

Microsoft Active Directory has an attribute, MaxValRange, which sets a limit on how many values for a multi-valued attribute will be returned. This is the range retrieval search extension. Essentially, this runs multiuple mini-searches, each returning a subset of the results within a given range, until all matches are returned.
For example, when doing a search for the member attribute, each entry could have multiple values, and there can be multiple entries with that attribute. If there are 2000 matching results (or more), then MaxValRange limits how many are displayed at once; this is the value range. The given attribute then has an additional flag set, showing which range in the set the result is in:
attribute:range=low-high:value
For example, results 100 to 500 in a search:
member;range=99-499: cn=John Smith...
This is described in the Microsoft documentation at http://msdn.microsoft.com/en-us/library/cc223242.aspx.
SSSD supports range retrievals with Active Directory providers as part of user and group management, without any additional configuration.
However, some LDAP provider attributes which are available to configure searches — such as ldap_user_search_base — are not performant with range retrievals. Be cautious when configuring search bases in the Active Directory provider domain and consider what searches may trigger a range retrieval.

12.2.13.3. Performance and LDAP Referrals

Referrals can negatively impact overall performance because of the time spent attempting to trace referrals. There is particularly bad performance degradation when referral chasing is used with an Active Directory identity provider. Disabling referral checking can significantly improve performance.
LDAP referrals are enabled by default, so they must be explicitly disabled in the LDAP domain configuration. For example:
ldap_referrals = false

12.2.13.4. Active Directory as Other Provider Types

Active Directory can be used as an identity provider and as an access, password, and authentication provider.
There are a number of options in the generic LDAP provider configuration which can be used to configure an Active Directory provider. Using the ad value is a short-cut which automatically pulls in the parameters and values to configure a given provider for Active Directory. For example, using access_provider = ad to configure an Active Directory access provider expands to this configuration using the explicit LDAP provider parameters:
access_provider = ldap
ldap_access_order = expire
ldap_account_expire_policy = ad

12.2.13.5. Configuring an Active Directory Identity Provider

Active Directory can work as a provider for identities, authentication, access control rules, and passwords, all of the *_provider parameters for a domain. Additionally, it is possible to load the native Active Directory schema for user and group entries, rather than using the default RFC 2307.
  1. Make sure that both the Active Directory and Linux systems have a properly configured environment.
    • Name resolution must be properly configured, particularly if service discovery is used with SSSD.
    • The clocks on both systems must be in sync for Kerberos to work properly.
  2. Set up the Linux system as an Active Directory client and enroll it within the Active Directory domain. This is done by configuring the Kerberos and Samba services on the Linux system.
    1. Set up Kerberos to use the Active Directory Kerberos realm.
      1. Open the Kerberos client configuration file.
        [root@server ~]# vim /etc/krb5.conf
      2. Configure the [logging] and [libdefaults] sections so that they connect to the Active Directory realm.
        [logging]
         default = FILE:/var/log/krb5libs.log
        
        [libdefaults]
         default_realm = EXAMPLE.COM
         dns_lookup_realm = true
         dns_lookup_kdc = true
         ticket_lifetime = 24h
         renew_lifetime = 7d
         rdns = false
         forwardable = yes
        If autodiscovery is not used with SSSD, then also configure the [realms] and [domain_realm] sections to explicitly define the Active Directory server.
    2. Configure the Samba server to connect to the Active directory server.
      1. Open the Samba configuration file.
        [root@server ~]# vim /etc/samba/smb.conf
      2. Set the Active Directory domain information in the [global] section.
        [global]
           workgroup = EXAMPLE
           client signing = yes
           client use spnego = yes
           kerberos method = secrets and keytab
           log file = /var/log/samba/%m.log
           password server = AD.EXAMPLE.COM
           realm = EXAMPLE.COM
           security = ads
    3. Add the Linux machine to the Active Directory domain.
      1. Obtain Kerberos credentials for a Windows administrative user.
        [root@server ~]# kinit Administrator
      2. Add the machine to the domain using the net command.
        [root@server ~]# net ads join -k
        Joined 'server' to dns domain 'example.com'
        This creates a new keytab file, /etc/krb5.keytab.
        List the keys for the system and check that the host principal is there.
        [root@server ~]# klist -k
  3. Use authconfig to enable SSSD for system authentication.
    # authconfig --update --enablesssd --enablesssdauth
  4. Set the Active Directory domain as an identity provider in the SSSD configuration, as shown in Example 12.7, “An Active Directory 2008 R2 Domain” and Example 12.8, “An Active Directory 2008 R2 Domain with ID Mapping”.
  5. Restart the SSH service to load the new PAM configuration.
    [root@server ~]# service restart sshd
  6. Restart SSSD after changing the configuration file.
    [root@rhel-server ~]# service sssd restart

Example 12.7. An Active Directory 2008 R2 Domain

[root@rhel-server ~]# vim /etc/sssd/sssd.conf

[sssd]
config_file_version = 2
domains = ad.example.com
services = nss, pam

...

[domain/ad.example.com]
id_provider = ad
ipa_server = ipaserver.example.com
ipa_hostname = ipa1.example.com
auth_provider = ad
chpass_provider = ad
access_provider = ad

# defines user/group schema type
ldap_schema = ad

# using explicit POSIX attributes in the Windows entries
ldap_id_mapping = False

# caching credentials
cache_credentials = true

# access controls
ldap_access_order = expire
ldap_account_expire_policy = ad
ldap_force_upper_case_realm = true

# performance
ldap_disable_referrals = true

There are two parameters that are critical for ID mapping: the Active Directory schema must be loaded (ldap_schema) and ID mapping must be explicitly enabled (ldap_id_mapping).

Example 12.8. An Active Directory 2008 R2 Domain with ID Mapping

[root@rhel-server ~]# vim /etc/sssd/sssd.conf

[sssd]
config_file_version = 2
domains = ad.example.com
services = nss, pam

...

[domain/ad.example.com]
id_provider = ad
ipa_server = ipaserver.example.com
ipa_hostname = ipa1.example.com
auth_provider = ad
chpass_provider = ad
access_provider = ad

# defines user/group schema type
ldap_schema = ad

# for SID-UID mapping
ldap_id_mapping = True

# caching credentials
cache_credentials = true

# access controls
ldap_access_order = expire
ldap_account_expire_policy = ad
ldap_force_upper_case_realm = true

# performance
ldap_disable_referrals = true

All of the potential configuration attributes for an Active Directory domain are listed in the sssd-ldap(5) and sssd-ad(5) man pages.

12.2.14. Configuring Domains: Active Directory as an LDAP Provider (Alternative)

While Active Directory can be configured as a type-specific identity provider, it can also be configured as a pure LDAP provider with a Kerberos authentication provider.
  1. It is recommended that SSSD connect to the Active Directory server using SASL, which means that the local host must have a service keytab for the Windows domain on the Linux host.
    This keytab can be created using Samba.
    1. Configure the /etc/krb5.conf file to use the Acti Directory realm.
      [logging]
       default = FILE:/var/log/krb5libs.log
      
      [libdefaults]
       default_realm = AD.EXAMPLE.COM
       dns_lookup_realm = true
       dns_lookup_kdc = true
       ticket_lifetime = 24h
       renew_lifetime = 7d
       rdns = false
       forwardable = yes
      
      [realms]
      # Define only if DNS lookups are not working
      # AD.EXAMPLE.COM = {
      #  kdc = server.ad.example.com
      #  admin_server = server.ad.example.com
      # }
      
      [domain_realm]
      # Define only if DNS lookups are not working
      # .ad.example.com = AD.EXAMPLE.COM
      # ad.example.com = AD.EXAMPLE.COM
    2. Set the Samba configuration file, /etc/samba/smb.conf, to point to the Windows Kerberos realm.
      [global]
         workgroup = EXAMPLE
         client signing = yes
         client use spnego = yes
         kerberos method = secrets and keytab
         log file = /var/log/samba/%m.log
         password server = AD.EXAMPLE.COM
         realm = EXAMPLE.COM
         security = ads
    3. Then, run the net ads command to log in as an administrator principal. This administrator account must have sufficient rights to add a machine to the Windows domain, but it does not require domain administrator privileges.
      [root@server ~]# net ads join -U Administrator
    4. Run net ads again to add the host machine to the domain. This can be done with the host principal (host/FQDN) or, optionally, with the NFS service (nfs/FQDN).
      [root@server ~]# net ads join createupn="host/rhel-server.example.com@AD.EXAMPLE.COM" -U Administrator
  2. Make sure that the Services for Unix package is installed on the Windows server.
  3. Set up the Windows domain which will be used with SSSD.
    1. On the Windows machine, open Server Manager.
    2. Create the Active Directory Domain Services role.
    3. Create a new domain, such as ad.example.com.
    4. Add the Identity Management for UNIX service to the Active Directory Domain Services role. Use the Unix NIS domain as the domain name in the configuration.
  4. On the Active Directory server, create a group for the Linux users.
    1. Open Administrative Tools and select Active Directory Users and Computers.
    2. Select the Active Directory domain, ad.example.com.
    3. In the Users tab, right-click and select Create a New Group.
    4. Name the new group unixusers, and save.
    5. Double-click the unixusers group entry, and open the Users tab.
    6. Open the Unix Attributes tab.
    7. Set the NIS domain to the NIS domain that was configured for ad.example.com and, optionally, set a group ID (GID) number.
  5. Configure a user to be part of the Unix group.
    1. Open Administrative Tools and select Active Directory Users and Computers.
    2. Select the Active Directory domain, ad.example.com.
    3. In the Users tab, right-click and select Create a New User.
    4. Name the new user aduser, and make sure that the User must change password at next logon and Lock account checkboxes are not selected.
      Then save the user.
    5. Double-click the aduser user entry, and open the Unix Attributes tab. Make sure that the Unix configuration matches that of the Active Directory domain and the unixgroup group:
      • The NIS domain, as created for the Active Directory domain
      • The UID
      • The login shell, to /bin/bash
      • The home directory, to /home/aduser
      • The primary group name, to unixusers

    TIP

    Password lookups on large directories can take several seconds per request. The initial user lookup is a call to the LDAP server. Unindexed searches are much more resource-intensive, and therefore take longer, than indexed searches because the server checks every entry in the directory for a match. To speed up user lookups, index the attributes that are searched for by SSSD:
    • uid
    • uidNumber
    • gidNumber
    • gecos
  6. On the Linux system, configure the SSSD domain.
    [root@rhel-server ~]# vim /etc/sssd/sssd.conf
    For a complete list of LDAP provider parameters, see the sssd-ldap(5) man pages.

    Example 12.9. An Active Directory 2008 R2 Domain with Services for Unix

    [sssd]
    config_file_version = 2
    domains = ad.example.com
    services = nss, pam
    
    ...
    
    [domain/ad.example.com]
    cache_credentials = true
    
    # for performance
    ldap_referrals = false
    
    id_provider = ldap
    auth_provider = krb5
    chpass_provider = krb5
    access_provider = ldap
    
    ldap_schema = rfc2307bis
    
    ldap_sasl_mech = GSSAPI
    ldap_sasl_authid = host/rhel-server.example.com@AD.EXAMPLE.COM 
    
    #provide the schema for services for unix
    ldap_schema = rfc2307bis
    
    ldap_user_search_base = ou=user accounts,dc=ad,dc=example,dc=com
    ldap_user_object_class = user
    ldap_user_home_directory = unixHomeDirectory
    ldap_user_principal = userPrincipalName
    
    # optional - set schema mapping
    # parameters are listed in sssd-ldap
    ldap_user_object_class = user
    ldap_user_name = sAMAccountName
    
    ldap_group_search_base = ou=groups,dc=ad,dc=example,dc=com
    ldap_group_object_class = group
    
    ldap_access_order = expire
    ldap_account_expire_policy = ad
    ldap_force_upper_case_realm = true
    ldap_disable_referrals = true
    
    krb5_realm = AD-REALM.EXAMPLE.COM
    # required
    krb5_canonicalize = false

  7. Restart SSSD.
    [root@rhel-server ~]# service sssd restart

12.2.15. Domain Options: Setting Username Formats

One of the primary actions that SSSD performs is mapping a local system user to an identity in the remote identity provider. SSSD uses a combination of the username and the domain backend name to create the login identity.
As long as they belong to different domains, SSSD can recognize different users with the same username. For example, SSSD can successfully authenticate both jsmith in the ldap.example.com domain and jsmith in the ldap.otherexample.com domain.
The name format used to construct full username is (optionally) defined universally in the [sssd] section of the configuration and can then be defined individually in each domain section.
Usernames for different services — LDAP, Samba, Active Directory, Identity Management, even the local system — all have different formats. The expression that SSSD uses to identiy username/domain name sets must be able to interpret names in different formats. This expression is set in the re_expression parameter.
In the global default, this filter constructs a name in the form name@domain:
(?P<name>[^@]+)@?(?P<domain>[^@]*$)

NOTE

The regular expression format is Python syntax.
The domain part may be supplied automatically, based on the domain name of the identity provider. Therefore, a user can log in as jsmith and if the user belongs to the LOCAL domain (for example), then his username is interpreted by SSSD as jsmith@LOCAL.
However, other identity providers may have other formats. Samba, for example, has a very strict format so that username must match the form DOMAIN\username. For Samba, then, the regular expression must be:
(?P<domain>[^\\]*?)\\?(?P<name>[^\\]+$)
Some providers, such as Active Directory, support multiple different name formats. Active Directory and Identity Management, for example, support three different formats by default:
  • username
  • username@domain.name
  • DOMAIN\username
The default value for Active Directory and Identity Management providers, then, is a more complex filter that allows all three name formats:
(((?P<domain>[^\\]+)\\(?P<name>.+$))|((?P<name>[^@]+)@(?P<domain>.+$))|(^(?P<name>[^@\\]+)$)) 

NOTE

Requesting information with the fully-qualified name, such as jsmith@ldap.example.com, always returns the proper user account. If there are multiple users with the same username in different domains, specifying only the username returns the user for whichever domain comes first in the lookup order.
While re_expression is the most important method for setting username formats, there are two other options which are useful for other applications.
Default Domain Name Value
The first sets a default domain name to be used with all users, default_domain_suffix. (This is a global setting, available in the [sssd] section only.) There may be a case where multiple domains are configured but only one stores user data and the others are used for host or service identities. Setting a default domain name allows users to log in with only their username, not specifying the domain name (which would be required for users outside the primary domain).
[sssd]
...
default_domain_suffix = USERS.EXAMPLE.COM
Full Name Format for Output
The other parameter is related to re_expression, only instead of defining how to interpret a username, it defines how to print an identified name. The full_name_format parameter sets how the username and domain name (once determined) are displayed.
full_name_format = %1$s@%2$s

12.2.16. Domain Options: Enabling Offline Authentication

User identities are always cached, as well as information about the domain services. However, user credentials are not cached by default. This means that SSSD always checks with the backend identity provider for authentication requests. If the identity provider is offline or unavailable, there is no way to process those authentication requests, so user authentication could fail.
It is possible to enable offline credentials caching, which stores credentials (after successful login) as part of the user account in the SSSD cache. Therefore, even if an identity provider is unavailable, users can still authenticate, using their stored credentials. Offline credentials caching is primarily configured in each individual domain entry, but there are some optional settings that can be set in the PAM service section, because credentials caching interacts with the local PAM service as well as the remote domain.
[domain/EXAMPLE]
cache_credentials = true
There are optional parameters that set when those credentials expire. Expiration is useful because it can prevent a user with a potentially outdated account or credentials from accessing local services indefinitely.
The credentials expiration itself is set in the PAM service, which processes authentication requests for the system.
[sssd]
services = nss,pam
...

[pam]
offline_credentials_expiration = 3
...

[domain/EXAMPLE]
cache_credentials = true
...
offline_credentials_expiration sets the number of days after a successful login that a single credentials entry for a user is preserved in cache. Setting this to zero (0) means that entries are kept forever.
While not related to the credentials cache specifically, each domain has configuration options on when individual user and service caches expire:
  • account_cache_expiration sets the number of days after a successful login that the entire user account entry is removed from the SSSD cache. This must be equal to or longer than the individual offline credentials cache expiration period.
  • entry_cache_timeout sets a validity period, in seconds, for all entries stored in the cache before SSSD requests updated information from the identity provider. There are also individual cache timeout parameters for group, service, netgroup, sudo, and autofs entries; these are listed in the sssd.conf man page. The default time is 5400 seconds (90 minutes).
For example:
[sssd]
services = nss,pam
...

[pam]
offline_credentials_expiration = 3
...

[domain/EXAMPLE]
cache_credentials = true
account_cache_expiration = 7
entry_cache_timeout = 14400
...

12.2.17. Domain Options: Setting Password Expirations

Password policies generally set an expiration time, when passwords expire and must be replaced. Those password expiration policies are evaluated by server-side, through the identity provider, and then a warning can be processed and displayed in SSSD through its PAM service.
There are two potential configuration areas for password warnings:
  • A global default for all domains on how far in advance of the password expiration to display a warning. This is set for the PAM service.
  • Per-domain settings on how far in advance of the password expiration to display a warning.
    When using a domain-level password expiration warning, an authentication provider (auth_provider) must also be configured for the domain.
For example:
[sssd]
services = nss,pam
...

[pam]
pam_pwd_expiration_warning = 3
...

[domain/EXAMPLE]
id_provider = ipa
auth_provider = ipa
pwd_expiration_warning = 7
The password expiration warning must be sent from the server to SSSD for the warning to be displayed. If no password warning is sent from the server, no message is displayed through SSSD, even if the password expiration time is within the period set in SSSD.
If the password expiration warning is not set in SSSD or is set to zero (0), then the SSSD password warning filter is not applied and the server-side password warning is automatically displayed.

NOTE

The PAM or domain password expirations essentially override (or ignore) the password warning settings on the backend identity provider — as long as the password warning is sent from the server.
For example, a backend identity provider has the warning set at 28 days, but the PAM service in SSSD has it set to seven days. The provider sends the warning to SSSD starting at 28 days, but the warning is not displayed locally until seven days, according to the password expiration set in the SSSD configuration.

TIP

A similar parameter is available when using Kerberos authentication providers to cache Kerberos credentials, krb5_store_password_if_offline.

12.2.18. Domain Options: Using DNS Service Discovery

DNS service discovery, defined in RFC 2782, allows applications to check the SRV records in a given domain for certain services of a certain type; it then returns any servers discovered of that type.
With SSSD, the identity and authentication providers can either be explicitly defined (by IP address or hostname) or they can be discovered dynamically, using service discovery. If no provider server is listed — for example, if id_provider = ldap is set without a corresponding ldap_uri parameter — then discovery is automatically used.
The DNS discovery query has this format:
_service._protocol.domain
For example, a scan for an LDAP server using TCP in the example.com domain looks like this:
_ldap._tcp.example.com

NOTE

For every service with which to use service discovery, add a special DNS record to the DNS server:
_service._protocol._domain TTL priority weight port hostname
For SSSD, the service type is LDAP by default, and almost all services use TCP (except for Kerberos, which starts with UDP). For service discovery to be enabled, the only thing that is required is the domain name. The default is to use the domain portion of the machine hostname, but another domain can be specified (using the dns_discovery_domain parameter).
So, by default, no additional configuration needs to be made for service discovery — with one exception. The password change provider has server discovery disabled by default, and it must be explicitly enabled by setting a service type.
[domain/EXAMPLE]
...
chpass_provider = ldap
ldap_chpass_dns_service_name = ldap
While no configuration is necessary, it is possible for server discovery to be customized by using a different DNS domain (dns_discovery_domain) or by setting a different service type to scan for. For example:
[domain/EXAMPLE]
id _provider = ldap

dns_discovery_domain = corp.example.com
ldap_dns_service_name = ldap

chpass_provider = krb5
ldap_chpass_dns_service_name = kerberos
Lastly, service discovery is never used with backup servers; it is only used for the primary server for a provider. What this means is that discovery can be used initially to locate a server, and then SSSD can fall back to using a backup server. To use discovery for the primary server, use _srv_ as the primary server value, and then list the backup servers. For example:
[domain/EXAMPLE]
id _provider = ldap
ldap_uri = _srv_
ldap_backup_uri = ldap://ldap2.example.com

auth_provider = krb5
krb5_server = _srv_
krb5_backup_server = kdc2.example.com

chpass_provider = krb5
ldap_chpass_dns_service_name = kerberos
ldap_chpass_uri = _srv_
ldap_chpass_backup_uri = kdc2.example.com

NOTE

Service discovery cannot be used with backup servers, only primary servers.
If a DNS lookup fails to return an IPv4 address for a hostname, SSSD attempts to look up an IPv6 address before returning a failure. This only ensures that the asynchronous resolver identifies the correct address.
The hostname resolution behavior is configured in the lookup family order option in the sssd.conf configuration file.

12.2.19. Domain Options: Using IP Addresses in Certificate Subject Names (LDAP Only)

Using an IP address in the ldap_uri option instead of the server name may cause the TLS/SSL connection to fail. TLS/SSL certificates contain the server name, not the IP address. However, the subject alternative name field in the certificate can be used to include the IP address of the server, which allows a successful secure connection using an IP address.
  1. Convert an existing certificate into a certificate request. The signing key (-signkey) is the key of the issuer of whatever CA originally issued the certificate. If this is done by an external CA, it requires a separate PEM file; if the certificate is self-signed, then this is the certificate itself. For example:
    openssl x509 -x509toreq -in old_cert.pem -out req.pem -signkey key.pem
    With a self-signed certificate:
    openssl x509 -x509toreq -in old_cert.pem -out req.pem -signkey old_cert.pem
  2. Edit the /etc/pki/tls/openssl.cnf configuration file to include the server's IP address under the [ v3_ca ] section:
    subjectAltName = IP:10.0.0.10
  3. Use the generated certificate request to generate a new self-signed certificate with the specified IP address:
    openssl x509 -req -in req.pem -out new_cert.pem -extfile ./openssl.cnf -extensions v3_ca -signkey old_cert.pem
    The -extensions option sets which extensions to use with the certificate. For this, it should be v3_ca to load the appropriate section.
  4. Copy the private key block from the old_cert.pem file into the new_cert.pem file to keep all relevant information in one file.
When creating a certificate through the certutil utility provided by the nss-utils package, note that certutil supports DNS subject alternative names for certificate creation only.

12.2.20. Creating Domains: Proxy

A proxy with SSSD is just a relay, an intermediary configuration. SSSD connects to its proxy service, and then that proxy loads the specified libraries. This allows SSSD to use some resources that it otherwise would not be able to use. For example, SSSD only supports LDAP and Kerberos as authentication providers, but using a proxy allows SSSD to use alternative authentication methods like a fingerprint scanner or smart card.

Table 12.9. Proxy Domain Configuration Parameters

Parameter Description
proxy_pam_target Specifies the target to which PAM must proxy as an authentication provider. The PAM target is a file containing PAM stack information in the default PAM directory, /etc/pam.d/.
This is used to proxy an authentication provider.

Important

Ensure that the proxy PAM stack does not recursively include pam_sss.so.
proxy_lib_name Specifies which existing NSS library to proxy identity requests through.
This is used to proxy an identity provider.

Example 12.10. Proxy Identity and Kerberos Authentication

The proxy library is loaded using the proxy_lib_name parameter. This library can be anything as long as it is compatible with the given authentication service. For a Kerberos authentication provider, it must be a Kerberos-compatible library, like NIS.
[domain/PROXY_KRB5]
auth_provider = krb5
krb5_server = kdc.example.com
krb5_realm = EXAMPLE.COM

id_provider = proxy
proxy_lib_name = nis
cache_credentials = true

Example 12.11. LDAP Identity and Proxy Authentication

The proxy library is loaded using the proxy_pam_target parameter. This library must be a PAM module that is compatible with the given identity provider. For example, this uses a PAM fingerprint module with LDAP:
[domain/LDAP_PROXY]
id_provider = ldap
ldap_uri = ldap://example.com
ldap_search_base = dc=example,dc=com

auth_provider = proxy
proxy_pam_target = sssdpamproxy
cache_credentials = true
After the SSSD domain is configured, make sure that the specified PAM files are configured. In this example, the target is sssdpamproxy, so create a /etc/pam.d/sssdpamproxy file and load the PAM/LDAP modules:
auth          required      pam_frprint.so
account       required      pam_frprint.so
password      required      pam_frprint.so
session       required      pam_frprint.so

Example 12.12. Proxy Identity and Authentication

SSSD can have a domain with both identity and authentication proxies. The only configuration given then are the proxy settings, proxy_pam_target for the authentication PAM module and proxy_lib_name for the service, like NIS or LDAP.
This example illustrates a possible configuration, but this is not a realistic configuration. If LDAP is used for identity and authentication, then both the identity and authentication providers should be set to the LDAP configuration, not a proxy.
[domain/PROXY_PROXY]
auth_provider = proxy
id_provider = proxy
proxy_lib_name = ldap
proxy_pam_target = sssdproxyldap
cache_credentials = true
Once the SSSD domain is added, then update the system settings to configure the proxy service:
  1. Create a /etc/pam.d/sssdproxyldap file which requires the pam_ldap.so module:
    auth          required      pam_ldap.so
    account       required      pam_ldap.so
    password      required      pam_ldap.so
    session       required      pam_ldap.so
  2. Make sure the nss-pam-ldap package is installed.
    [root@server ~]# yum install nss-pam-ldap
  3. Edit the /etc/nslcd.conf file, the configuration file for the LDAP name service daemon, to contain the information for the LDAP directory:
    uid nslcd
    gid ldap
    uri ldaps://ldap.example.com:636
    base dc=example,dc=com
    ssl on
    tls_cacertdir /etc/openldap/cacerts

12.2.21. Creating Domains: Kerberos Authentication

Both LDAP and proxy identity providers can use a separate Kerberos domain to supply authentication. Configuring a Kerberos authentication provider requires the key distribution center (KDC) and the Kerberos domain. All of the principal names must be available in the specified identity provider; if they are not, SSSD constructs the principals using the format username@REALM.

Note

Kerberos can only provide authentication; it cannot provide an identity database.
SSSD assumes that the Kerberos KDC is also a Kerberos kadmin server. However, production environments commonly have multiple, read-only replicas of the KDC and only a single kadmin server. Use the krb5_kpasswd option to specify where the password changing service is running or if it is running on a non-default port. If the krb5_kpasswd option is not defined, SSSD tries to use the Kerberos KDC to change the password.
The basic Kerberos configuration options are listed in Table 12.10, “Kerberos Authentication Configuration Parameters”. The sssd-krb5(5) man page has more information about Kerberos configuration options.

Example 12.13. Basic Kerberos Authentication

# A domain with identities provided by LDAP and authentication by Kerberos
[domain/KRBDOMAIN]
id_provider = ldap
chpass_provider = krb5
ldap_uri = ldap://ldap.example.com
ldap_search_base = dc=example,dc=com
ldap-tls_reqcert = demand
ldap_tls_cacert = /etc/pki/tls/certs/ca-bundle.crt

auth_provider = krb5
krb5_server = kdc.example.com
krb5_backup_server = kerberos.example.com
krb5_realm = EXAMPLE.COM
krb5_kpasswd = kerberos.admin.example.com
krb5_auth_timeout = 15

Example 12.14. Setting Kerberos Ticket Renewal Options

The Kerberos authentication provider, among other tasks, requests ticket granting tickets (TGT) for users and services. These tickets are used to generate other tickets dynamically for specific services, as accessed by the ticket principal (the user).
The TGT initially granted to the user principal is valid only for the lifetime of the ticket (by default, whatever is configured in the configured KDC). After that, the ticket cannot be renewed or extended. However, not renewing tickets can cause problems with some services when they try to access a service in the middle of operations and their ticket has expired.
Kerberos tickets are not renewable by default, but ticket renewal can be enabled using the krb5_renewable_lifetime and krb5_renew_interval parameters.
The lifetime for a ticket is set in SSSD with the krb5_lifetime parameter. This specifies how long a single ticket is valid, and overrides any values in the KDC.
Ticket renewal itself is enabled in the krb5_renewable_lifetime parameter, which sets the maximum lifetime of the ticket, counting all renewals.
For example, the ticket lifetime is set at one hour and the renewable lifetime is set at 24 hours:
krb5_lifetime = 1h
krb5_renewable_lifetime = 1d
This means that the ticket expires every hour and can be renewed continually up to one day.
The lifetime and renewable lifetime values can be in seconds (s), minutes (m), hours (h), or days (d).
The other option — which must also be set for ticket renewal — is the krb5_renew_interval parameter, which sets how frequently SSSD checks to see if the ticket needs to be renewed. At half of the ticket lifetime (whatever that setting is), the ticket is renewed automatically. (This value is always in seconds.)
krb5_lifetime = 1h
krb5_renewable_lifetime = 1d
krb5_renew_interval = 60s

NOTE

If the krb5_renewable_lifetime value is not set or the krb5_renew_interval parameter is not set or is set to zero (0), then ticket renewal is disabled. Both krb5_renewable_lifetime and krb5_renew_interval are required for ticket renewal to be enabled.

Table 12.10. Kerberos Authentication Configuration Parameters

Parameter Description
chpass_provider Specifies which service to use for password change operations. This is assumed to be the same as the authentication provider. To use Kerberos, set this to krb5.
krb5_server Gives the primary Kerberos server, by IP address or hostnames, to which SSSD will connect.
krb5_backup_server Gives a comma-separated list of IP addresses or hostnames of Kerberos servers to which SSSD will connect if the primary server is not available. The list is given in order of preference, so the first server in the list is tried first.
After an hour, SSSD will attempt to reconnect to the primary service specified in the krb5_server parameter.
When using service discovery for KDC or kpasswd servers, SSSD first searches for DNS entries that specify UDP as the connection protocol, and then falls back to TCP.
krb5_realm Identies the Kerberos realm served by the KDC.
krb5_lifetime Requests a Kerberos ticket with the specified lifetime in seconds (s), minutes (m), hours (h) or days (d).
krb5_renewable_lifetime Requests a renewable Kerberos ticket with a total lifetime that is specified in seconds (s), minutes (m), hours (h) or days (d).
krb5_renew_interval Sets the time, in seconds, for SSSD to check if tickets should be renewed. Tickets are renewed automatically once they exceed half their lifetime. If this option is missing or set to zero, then automatic ticket renewal is disabled.
krb5_store_password_if_offline Sets whether to store user passwords if the Kerberos authentication provider is offline, and then to use that cache to request tickets when the provider is back online. The default is false, which does not store passwords.
krb5_kpasswd Lists alternate Kerberos kadmin servers to use if the change password service is not running on the KDC.
krb5_ccname_template Gives the directory to use to store the user's credential cache. This can be templatized, and the following tokens are supported:
  • %u, the user's login name
  • %U, the user's login UID
  • %p, the user's principal name
  • %r, the realm name
  • %h, the user's home directory
  • %d, the value of the krb5ccache_dir parameter
  • %P, the process ID of the SSSD client.
  • %%, a literal percent sign (%)
  • XXXXXX, a string at the end of the template which instructs SSSD to create a unique filename safely
For example:
krb5_ccname_template = FILE:%d/krb5cc_%U_XXXXXX
krb5_ccachedir Specifies the directory to store credential caches. This can be templatized, using the same tokens as krb5_ccname_template, except for %d and %P. If %u, %U, %p, or %h are used, then SSSD creates a private directory for each user; otherwise, it creates a public directory.
krb5_auth_timeout Gives the time, in seconds, before an online authentication or change password request is aborted. If possible, the authentication request is continued offline. The default is 15 seconds.

12.2.22. Creating Domains: Access Control

SSSD provides a rudimentary access control for domain configuration, allowing either simple user allow/deny lists or using the LDAP backend itself.

12.2.22.1. Using the Simple Access Provider

The Simple Access Provider allows or denies access based on a list of usernames or groups.
The Simple Access Provider is a way to restrict access to certain, specific machines. For example, if a company uses laptops, the Simple Access Provider can be used to restrict access to only a specific user or a specific group, even if a different user authenticated successfully against the same authentication provider.
The most common options are simple_allow_users and simple_allow_groups, which grant access explicitly to specific users (either the given users or group members) and deny access to everyone else. It is also possible to create deny lists (which deny access only to explicit people and implicitly allow everyone else access).
The Simple Access Provider adheres to the following four rules to determine which users should or should not be granted access:
  • If both the allow and deny lists are empty, access is granted.
  • If any list is provided, allow rules are evaluated first, and then deny rules. Practically, this means that deny rules supersede allow rules.
  • If an allowed list is provided, then all users are denied access unless they are in the list.
  • If only deny lists are provided, then all users are allowed access unless they are in the list.
This example grants access to two users and anyone who belongs to the IT group; implicitly, all other users are denied:
[domain/example.com]
access_provider = simple
simple_allow_users = jsmith,bjensen
simple_allow_groups = itgroup

Note

The LOCAL domain in SSSD does not support simple as an access provider.
Other options are listed in the sssd-simple man page, but these are rarely used.