Red Hat Training

A Red Hat training course is available for Red Hat Enterprise Linux

Hypervisor Deployment Guide

Red Hat Enterprise Linux 5

The complete guide to installing, deploying and maintaining Red Hat Enterprise Virtualization Hypervisors.

Edition 9

Chris Curran

Red Hat Engineering Content Services

David Jorm

Red Hat Engineering Content Services

Stephen Gordon

Red Hat Engineering Content Services

Laura Bailey

Red Hat Engineering Content Services

Abstract

This book is a guide to installing and administering Red Hat Enterprise Virtualization Hypervisor.

Preface

This book is a guide to installing Red Hat Enterprise Virtualization (RHEV) Hypervisors. This book provides a step-by-step procedures for installing Hypervisors with Red Hat Enterprise Virtualization Manager for Desktops. Alternative and advanced installation and configuration options are covered to assist users with varied environments and systems.

Chapter 1. Introducing Red Hat Enterprise Virtualization Hypervisors

The Red Hat Enterprise Virtualization (RHEV) Hypervisor is a compact, full-featured virtualization platform for quickly and easily deploying and managing virtual machines. The RHEV Hypervisor is part of the Red Hat Enterprise Virtualization suite. The RHEV Hypervisor is designed for integration with the Red Hat Enterprise Virtualization Manager for Servers and the Red Hat Enterprise Virtualization Manager for Desktops.
The RHEV Hypervisor can be installed from USB storage devices, CD-ROMs, DVDs, pre-installed by an OEM or provisioned with PXE.
The RHEV Hypervisor is based on the Kernel-based Virtual Machine (KVM). KVM is an advanced and efficient virtualization hypervisor implemented as a Linux kernel module. As KVM is a kernel module, it leverages the existing Red Hat Enterprise Linux kernel and benefits from the default kernel's extensive testing, device support and flexibility.
The RHEV Hypervisor is constructed using a subset of packages from Red Hat Enterprise Linux along with several special RHEV packages. RHEV is certified for all hardware certified for Red Hat Enterprise Linux (except where noted, refer to Chapter 3, Requirements and limitations of Red Hat Enterprise Virtualization Hypervisors).

Chapter 2. Understanding the Red Hat Enterprise Virtualization Hypervisor

This chapter provides a brief introduction to the technical aspects of the Red Hat Enterprise Virtualization Hypervisor. This chapter is written to assist technical users with identifying issues and understanding the Red Hat Enterprise Virtualization Hypervisor.
The Red Hat Enterprise Virtualization Hypervisor is derived from the Managed Node component of the oVirt project. While the core technology of oVirt Managed Node is used, the Red Hat Enterprise Virtualization Hypervisor incorporates its own management communications stack. For more information on the oVirt project, visit oVirt.org.

2.1. Red Hat Enterprise Virtualization Hypervisor directories

Red Hat Enterprise Virtualization Hypervisors have a directory layout that varies from Red Hat Enterprise Linux and other Linux Distributions. All Red Hat Enterprise Hypervisors have the following directories:
/config
The /config directory contains all persistent configuration files for the Red Hat Enterprise Virtualization Hypervisor. These files control passwords, storage configuration, security and networking.
The /config directory must be at least 5MB.
/boot
The /boot directory contains the boot loader, the kernel and the initramfs file.
The /boot directory is not mounted by default. The ovirt-config-boot boot command mounts the partition only when needed for an upgrade.
The /boot directory must be exactly 50MB.
/liveos
The /liveos directory contains a compressed Red Hat Enterprise Virtualization Hypervisor live CD image. The Red Hat Enterprise Virtualization Hypervisor boots and runs from the ISO image file in this directory.
The /liveos directory is not normally visible on the running system. This is the folder containing the livecd ISO. During an upgrade /dev/HostVG/Root is temporarily mounted to /liveos
/var/log
Contains all the logs for the Hypervisor.
The log directory must be at least 2048MB. The default size of the log directory is 2048MB.
/var/log/core
Contains core dumps from the Hypervisor which can be used for debugging and support.
/var/vdsm/
The /var/vdsm/ is used by the vdsmd daemon for storing volatile data, including /var/vdsm/ts which stores vdsm truststore.
/rhev/data-center
Contains links to Storage Domains.
/data
This directory contains virtual machine cache data and other miscellaneous files.
The data partition must be at least as large as the RAM on the host system plus an additional 512MB in size. A data partition at least one and a half times as large as the RAM on the host system is recommended.
The default size for the data partition is the remaining available disk space.

Chapter 3. Requirements and limitations of Red Hat Enterprise Virtualization Hypervisors

This chapter contains all system requirements and limitations which apply to Red Hat Enterprise Virtualization (RHEV) Hypervisors. These requirements are determined by present hardware and software limits as well as testing and support considerations. System requirements and limitations will vary over time due to ongoing software development and hardware improvements.

3.1. System requirements

Red Hat Enterprise Virtualization Hypervisors have the following hardware requirements:
  • A CPU with the hardware virtualization extensions; either an AMD system with AMD-V™ or an Intel system with Intel VT®.
    When the Red Hat Enterprise Virtualization Hypervisor boots a message may appear:
    Virtualization hardware is unavailable.
       (No virtualization hardware was detected on this system)
    This warning indicates the virtualization extensions are either disabled or not present on your processor.

    Important

    Some systems have the virtualization extensions disabled in BIOS; virtualization extensions must be enabled in BIOS. First, boot into rescue mode (by typing linux rescue at the RHEV Hypervisor boot prompt) or by entering the shell from the configuration menu. Determine whether your processor has virtualization extensions with this command:
    $ grep -E 'svm|vmx' /proc/cpuinfo
    If any output is shown, the processor is hardware virtualization capable. However in some circumstances manufacturers disable the virtualization extensions in BIOS. As an additional check, verify that the kvm modules are loaded in the kernel:
    # lsmod | grep kvm
    If the output includes kvm_intel or kvm_amd then the kvm hardware virtualization modules are loaded and your system meets requirements.
  • An Intel® 64 or AMD64 CPU extension.
  • At least one network controller with a minimum bandwidth of 1Gbps.
  • At least 512MB of RAM.
  • Sufficient RAM for guest virtual machines. The amount of RAM required for virtual machines depends on:
    • the guest operating system's requirements,
    • the virtual machine's application requirements, and
    • memory activity and usage of the virtual machine.
    KVM can overcommit physical RAM for virtual machines by only allocating RAM for virtual machines as required, and shifting under-utilized virtual machines into swap.
  • The minimum supported internal storage for a Hypervisor is the total of the following list:
    • The root partitions require exactly 512MB of storage.
    • The boot partition requires exactly 50MB of storage.
    • The logging partition is required and the recommended size is 2048MB.
    • The default and minimum size for the configuration partition is 5MB.
    • Swap is required and the recommended size is the recommended swap partition (from this article on Red Hat Knowledgebase) and the overcommit ratio multiplied by the RAM (the default overcommit ratio is 0.5). The minimum size for the swap partition is 5 MB.
      Recommended swap + (RAM * overcommit) = swap partition size
    • The default size of the data partition is the size of the system's RAM plus 512MB. The minimum size is 256MB.

Example 3.1. Determining minimum storage requirements

This example covers the minimum required storage for a RHEV Hypervisor installation on a system with 14GB of RAM (approximately 14,336MB).
The root partitions          = 512MB
The boot partition          = 50MB
The logging partition       = 2048MB
The config partition        = 5MB
The swap partition (using a 0.6 overcommit ratio)
4GB (4096MB) + (0.6 * 14GB) = 12698MB
The data partition is the size of the system's RAM plus 512MB
             (14GB + 512MB) = 14750MB
-------------------------------------------------------------
                      Total = 30063MB
Note, that this is for the minimum storage for an installation. It is recommended to use the defaults, which will use more storage space.

3.2. Limitations

The following limitations apply to hardware the Hypervisor is supported on and guest virtual machines:
  • A maximum of 64 physical CPUs on the host.
  • A maximum of 1TB of RAM.
  • A maximum of 16 virtualized CPUs per virtual machine.
  • A maximum of 256GB of virtualized RAM per 64 bit virtual machine.
  • A maximum of 4GB of virtualized RAM per 32 bit virtual machine. Note, the virtual machine may not register the entire 4GB. How much RAM the virtual machine recognizes is limited by its operating system.
  • A maximum of 8 virtualized storage devices per virtual machine.
  • A maximum of 8 virtualized network interface controllers per virtual machine.
  • A maximum of 32 virtualized PCI devices per virtual machine.

3.3. RHEV guest operating system support

RHEV presently supports the following operating systems for use on its guest virtual machines:
  • Red Hat Enterprise Linux 3 (32 bit and 64 bit)
  • Red Hat Enterprise Linux 4 (32 bit and 64 bit)
  • Red Hat Enterprise Linux 5 (32 bit and 64 bit)
  • Windows XP Service Pack 3 and newer (32 bit only)
  • Windows Server 2003 Service Pack 2 and newer (32 bit and 64 bit)
  • Windows Server 2008 (32 bit and 64 bit)
  • Windows Server 2008 R2 (64 bit only)
  • Windows 7 (32 bit and 64 bit)

3.4. Para-virtualized driver support

The para-virtualized drivers (the virtio drivers) support the following guest operating systems and versions. The para-virtualized drivers increase the performance for a virtual machine's block and network devices.
Guest operating systemPara-virtualized drivers
Windows XPBlock and network drivers
Windows 7 (32 bit and 64 bit)Block and network drivers
Windows Server 2008 (32 bit and 64 bit)Block and network drivers
Windows Server 2003 R2 (32 bit and 64 bit)Block and network drivers
Red Hat Enterprise Linux 4.8 and newer (32 bit and 64 bit)Block and network drivers
Red Hat Enterprise Linux 5.4 and newer (32 bit and 64 bit)Block and network drivers

Chapter 4. Preparing Red Hat Enterprise Virtualization Hypervisor installation media

This chapter covers creating installation media and preparing your systems before installing a Red Hat Enterprise Virtualization (RHEV) Hypervisor.
This chapter covers installing RHEV Hypervisors on a local storage device. This storage device can be a removable USB storage device, or an internal hard disk drive or solid state drive. Once the RHEV Hypervisor is installed, the system will boot the RHEV Hypervisor and all configuration data is preserved on the system.

4.1. Preparation instructions

The rhev-hypervisor package is needed for installation of RHEV Hypervisors. The rhev-hypervisor package contains the RHEV Hypervisor CD-ROM image. The following procedure installs the rhev-hypervisor package.
Entitlements to the Red Hat Enterprise Virtualization Hypervisor 5 channel must be available to your RHN account to download the RHEV Hypervisor images.

Procedure 4.1. Downloading and Installing the Package

The rhev-hypervisor* package contains the Red Hat Enterprise Virtualization Hypervisor ISO image. The ISO itself contains additional tools for USB and PXE installations.
  1. Download the latest rhev-hypervisor* package from Red Hat Network onto a Red Hat Enterprise Linux system to which you have root access. The list of Hypervisor packages is available at https://rhn.redhat.com/rhn/channels/PackageList.do?cid=9462.
  2. As root, navigate to the location of the downloaded package. Install it to the system in preparation for creation of the boot media.
    # yum localinstall rhev-hypervisor*.rpm
  3. Extract the livecd-iso-to-pxeboot and livecd-iso-to-disk commands from the ISO. To do this:
    1. Change to the directory the Hypervisor ISO is installed in, while logged in as root.
      # cd /usr/share/rhev-hypervisor
    2. Create a directory to which the ISO file will be mounted on a temporary basis.
      # mkdir iso/
    3. Mount the ISO file as a loopback device. Use the directory created in the previous step as the mount target.
      # mount -o loop rhev-hypervisor.iso iso/
    4. Copy the livecd-iso-to-pxeboot and livecd-iso-to-disk scripts to the current directory.
      # cp iso/LiveOS/livecd-iso-to-* ./
    5. Unmount the ISO file and remove the temporary directory.
      # umount iso/
      # rmdir iso/
The RHEV Hypervisor ISO image is located in the /usr/share/rhev-hypervisor/ directory and named rhev-hypervisor.iso.

4.1.1. BIOS settings and boot process troubleshooting

Before installing RHEV Hypervisors you should verify your BIOS is correctly configured for the installation method you intend to use. Many motherboard and PC manufacturers disable different booting methods in the BIOS. Most BIOS chips boot from the following devices in order:
  1. 3.5 inch diskette
  2. CD-ROM or DVD device
  3. Local hard disk
Many BIOS chips have disabled one or more of the following boot methods: USB storage devices, CD-ROMs, DVDs or network boot. To boot from your chosen method, enable the method or device and set that device as the first boot device in BIOS.
Not all motherboards support the boot methods described in this chapter. Consult your motherboard or system manufacturer's documentation on whether it is possible to use a particular boot method. That said, many modern systems support all of the boot methods listed in this chapter.

Warning

BIOS settings vary from manufacturer to manufacturer. Therefore, any examples of settings may be inaccurate for some systems. Due to this inconsistency, you should review your motherboard or system manufacturer's documentation.

4.1.2. Confirm hardware virtualization support

Verify that your system supports RHEV Hypervisors. RHEV Hypervisors require that the virtualization extensions are present and enabled in the BIOS before the Hypervisor is installed.
  1. Boot the RHEV Hypervisor from removable media. For example, a USB stick or CD-ROM.
  2. Once the Hypervisor boot prompt is displayed, enter the command:
    : linux rescue
  3. Once the Hypervisor boots, verify your CPU contains the virtualization extensions with the following command:
    # grep -E 'svm|vmx' /proc/cpuinfo
    Output displays if the processor has the hardware virtualization extensions.
  4. Verify that the KVM modules load by default:
    # lsmod | grep kvm
    If the output includes kvm_intel or kvm_amd then the kvm hardware virtualization modules are loaded and the system meets the requirements.

4.2. Deploying RHEV Hypervisors with PXE and tftp

This section covers installing RHEV Hypervisors over a network with PXE and tftp. Configuring the DHCP and tftp servers for PXE booting is outside the scope of this book.
  1. Install the rhev-hypervisor package and extract the livecd-iso-to-pxeboot script. Refer to Procedure 4.1, “Downloading and Installing the Package”.
  2. Create vmlinuz and initrd images with livecd-iso-to-pxeboot:
    # /usr/share/rhev-hypervisor/livecd-iso-to-pxeboot /usr/share/rhev-hypervisor/rhev-hypervisor.iso
    Your pxeboot image is complete.
    
    Copy tftpboot/ subdirectory to /tftpboot or a subdirectory of /tftpboot.
    Set up your DHCP, TFTP and PXE server to serve /tftpboot/.../pxeboot.0
    
    Note: The initrd image contains the whole CD ISO and is consequently
    very large.  You will notice when pxebooting that initrd can take a
    long time to download.  This is normal behaviour.
  3. The output of livecd-iso-to-pxeboot command is a directory called tftpboot that has the following files in it:
    • pxelinux.0
    • pxelinux.cfg/default
    • vmlinuz0
    • initrd0.img
  4. The vmlinuz and initrd files can be imported into a cobbler server or any other PXE and tftp servers.
    Import the files to the appropriate directory.
  5. The pxelinux.cfg/default file provides a template for configuring the PXE server to export the Hypervisor image:
    DEFAULT pxeboot
    TIMEOUT 20
    PROMPT 0
    LABEL pxeboot
            KERNEL vmlinuz0
            APPEND rootflags=loop initrd=initrd0.img root=/rhev-hypervisor.iso rootfstype=auto ro liveimg
                   check crashkernel=128M@16M elevator=deadline processor.max_cstate=1
    ONERROR LOCALBOOT 0
    PXE booted Hypervisors rely on the PXE server passing the MAC address of the PXE interface to the kernel. This is provided by using the IPAPPEND 2 parameter.
    Modify the templates as required for your environment.

Warning

The root=/rhev-hypervisor.iso parameter in pxelinux.cfg/default is a default value. If the ISO file you are using has a name other than rhev-hypervisor.iso it should be passed when calling livecd-iso-to-pxeboot. For example, for the ISO file rhev_hypervisor_2_2.iso use the command livecd-iso-to-pxeboot rhev_hypervisor_2_2.iso. This will produce the correct parameter root=/rhev_hypervisor_2_2.iso in pxelinux.cfg/default.

4.2.1. Booting a RHEV Hypervisor with PXE

For network booting the network interface card must support PXE booting.
To boot a RHEV Hypervisor from a PXE server:
  1. Enter your system's BIOS. On most systems, the key or combination of keys is prompted shortly after the system has power. Usually, this key is delete, F1 or F2.
  2. Enable network booting if network booting is disabled.
  3. Set the network interface card as the first boot device.
  4. Boot the system. If the PXE parameters are configured correctly an automated installation should begin.
  5. Change or disable network booting after the RHEV Hypervisor is installed. This is to avoid overwriting the installation on each reboot (unless this is desired functionality) and to prevent certain security vulnerabilities.

Warning

The network interface used for PXE boot installation must be same interface used to connect to the RHEV Manager.

Note

For more information on the kernel parameters, refer to Section 5.2, “RHEV Hypervisor Kernel Parameters and Automated Installation”.
The RHEV Hypervisor should now be installed.

4.3. Preparing a RHEV Hypervisor USB storage device

RHEV Hypervisors can install from USB storage devices. They can also be installed to USB storage devices or solid state disks. However, the initial boot/install USB device must be a separate device from the installation target. Network booting with PXE and tftp provides the greatest flexibility and scalability. For environments where network restrictions prevent network booting, or for systems without PXE capable network interface cards, a local media installation such as CD-ROM or USB is necessary. Booting from USB storage devices is a useful alternative to booting from CD, for systems without CD-ROM drives.

Note

Not all systems support booting from a USB storage device. Ensure that your system's BIOS supports booting from USB storage devices before proceeding.

4.3.1. Making a bootable RHEV Hypervisor USB storage device

This section covers making bootable RHEV Hypervisors on USB storage devices.
The livecd-iso-to-disk command can install a RHEV Hypervisor onto a USB storage device. The livecd-iso-to-disk command is part of the rhev-hypervisor package. Devices created with this command can boot the RHEV Hypervisors on systems which support booting via USB.
The basic livecd-iso-to-disk command usage follows this structure:
# /usr/share/rhev-hypervisor/livecd-iso-to-disk image device
Where the device parameter is the partition name of the USB storage device to install to. The image parameter is a ISO image of the RHEV Hypervisor. The default RHEV Hypervisor image location is /usr/share/rhev-hypervisor/rhev-hypervisor.iso. The livecd-iso-to-disk command requires devices to be formatted with the FAT or EXT3 file system.

Note

livecd-iso-to-disk uses a FAT or EXT3 formatted partition or block device.
USB storage devices are sometimes formatted without a partition table, use /dev/sdb or similar device name.
When a USB storage device is formatted with a partition table, use /dev/sdb1 or similar device name.
  1. Install the rhev-hypervisor package and extract the livecd-iso-to-disk script. Refer to Procedure 4.1, “Downloading and Installing the Package”.
  2. Use the livecd-iso-to-disk command to copy the .iso file to the disk. The --format parameter formats the disk. The --reset-mbr initializes the Master Boot Record (MBR). The example uses a USB storage device named /dev/sdc.
    # /usr/share/rhev-hypervisor/livecd-iso-to-disk --format --reset-mbr /usr/share/rhev-hypervisor/rhev-hypervisor.iso /dev/sdc
    Verifying image...
    /usr/share/rhev-hypervisor/rhev-hypervisor.iso:   eccc12a0530b9f22e5ba62b848922309
    Fragment sums: 8688f5473e9c176a73f7a37499358557e6c397c9ce2dafb5eca5498fb586
    Fragment count: 20
    Checking: 100.0%
    
    The media check is complete, the result is: PASS.
    
    It is OK to use this media.
    Copying live image to USB stick
    Updating boot config file
    Installing boot loader
    syslinux: only 512-byte sectors are supported
    USB stick set up as live image!
The USB storage device (/dev/sdc1) is ready to boot a RHEV Hypervisor.

4.3.2. Booting a RHEV Hypervisor USB storage device

Booting a RHEV Hypervisor from a USB storage device is similar to booting other live USB operating systems. To boot from a USB storage device:
  1. Enter the system's BIOS menu to enable USB storage device booting if not already enabled.
    1. Enable USB booting if this feature is disabled.
    2. Set booting USB storage devices to be first boot device.
    3. Shut down the system.
  2. Insert the RHEV Hypervisor bootable USB storage device.
  3. Restart the system
  4. The RHEV Hypervisor should boot automatically.
If the Hypervisor is running, you must now initialize the local storage device. Refer to Section 5.1.1, “Booting the Hypervisor for installation” for details.

4.4. Preparing a RHEV Hypervisor from a CD-ROM or DVD

RHEV Hypervisors can be installed with a CD-ROM or DVD.

4.4.1. Making a bootable RHEV Hypervisor CD-ROM

Burn the RHEV Hypervisor image to a CD-ROM with the cdrecord command. The cdrecord command is part of the cdrecord package which is installed on Red Hat Enterprise Linux by default.
  1. Verify that the cdrecord package is installed on the system.
    # rpm -q cdrecord
    cdrecord-2.01-10.7.el5
    If the package version is in the output, as seen above, the package is available.
    If it is not present, install cdrecord:
    # yum install cdrecord
  2. Insert a blank CD-ROM or DVD into your CD or DVD writer.
  3. Record the ISO file to the disc. The cdrecord command uses the following:
    cdrecord dev=device /iso/file/path/
    This example uses the first CD-RW (/dev/cdrw) device available and the default RHEV Hypervisor image location, /usr/share/rhev-hypervisor/rhev-hypervisor.iso.
    # cdrecord dev=/dev/cdrw /usr/share/rhev-hypervisor/rhev-hypervisor.iso
If no errors occurred, the RHEV Hypervisor is ready to boot. Errors sometimes occur during the recording process due to errors on the media itself. If this occurs insert another writable disk and repeat the command above.
The RHEV Hypervisor uses a program (isomd5sum) to verify the integrity of the installation media every time the Hypervisor is booted. If media errors are reported in the boot sequence you have a bad CD-ROM. Follow the procedure above to create a new CD-ROM or DVD.

4.4.2. Booting a RHEV Hypervisor CD-ROM

For many systems, the default BIOS configuration boots from CD-ROM first. If booting from CD-ROM is disabled or is not the first boot device refer to Section 4.1.1, “BIOS settings and boot process troubleshooting” and your manufacturers manuals for more information.
To boot from CD-ROM, insert the RHEV Hypervisor CD-ROM and turn the computer on.
The RHEV Hypervisor should start to boot. If the RHEV Hypervisor does not start to boot your BIOS may not be configured to boot from CD-ROM first or booting from CD-ROM may be disabled.
If the Hypervisor is running, you must now initialize the local storage device. Refer to Section 5.1.1, “Booting the Hypervisor for installation” for details.

Chapter 5. Installing Red Hat Enterprise Virtualization Hypervisors

This chapter covers installing Red Hat Enterprise Virtualization (RHEV) Hypervisors.
Red Hat Enterprise Virtualization requires a RHEV Manager and one or more RHEV Hypervisors or Red Hat Enterprise Linux hosts. Installing the Red Hat Enterprise Virtualization Manager and Red Hat Enterprise Linux hosts is covered by the Red Hat Enterprise Virtualization Manager for Servers Installation Guide. This chapter covers installing Red Hat Enterprise Virtualization Hypervisors.
RHEV Hypervisors can use Storage Area Networks (SANs) and other network storage for storing virtual machine images. RHEV Hypervisors can be installed on SANs, provided that the Host Bus Adaptor (HBA) can be configured as a boot device in BIOS.
RHEV Hypervisors can use multipath devices for installation. Multipath is often used for SANs or other networked storage. Multipath is enabled by default at install time. Any block device which responds to scsi_id can use multipath. This is most devices, except USB storage and some older ATA disks.

Important

The RHEV Manager must be installed before RHEV Hypervisors are installed. Refer to the Red Hat Enterprise Virtualization for Servers Installation Guide for instructions on installing the RHEV Manager.
There are two methods for installing RHEV Hypervisors:

5.1. Red Hat Enterprise Virtualization Hypervisor interactive installation

RHEV Hypervisors must be installed on physical servers, not virtual machines.
The instructions in this section can be used for installing on multiple systems. Always remember to use unique hostnames and IP addresses to avoid networking conflicts.

5.1.1. Booting the Hypervisor for installation

There are several methods for booting Hypervisors, refer to Chapter 4, Preparing Red Hat Enterprise Virtualization Hypervisor installation media for detailed instructions on preparing boot media for RHEV Hypervisor installation.

Procedure 5.1. Booting the Hypervisor

  1. Insert the RHEV Hypervisor installation media.
  2. Power on the system and ensure the system boots from the installation media.
  3. Once the RHEV Hypervisor has completed the boot phase, the kernel parameter prompt appears.
    ISOLINUX 3.11 2005-09-02  Copyright (C) 1994-2005 H. Peter Anvin
    This is the Red Hat Enterprise Virtualization Hypervisor 5.7-2.2 (4.2)
    - to boot press the <ENTER> key
    or type linux and press the <ENTER> key.
    - Extra kernel command line parameters can be passed following the 
    linux keyword.
    - Boot parameters summary:
    BOOTIF=link|eth*|<MAC>(pxelinux format)
    storage_init=[usb|scsi|cciss[:serial#]]|/dev/...
    storage_vol=:SWAP_MB::CONFIG_MB:LOGGING_MB:DATA_MB
    mem_overcommit=<overcommit_ratio>
    upgrade
    ip=<client-ip> netmask=<netmask> gateway=<gw-ip>
    vlan=<VLAN-id>
    dns=server[:server]
    ntp=server[:server]
    hostname=fqdn
    management_server=RHEVM-host[:port]
    netconsole_server=RHEVM-host[:port]
    boot:
    Enter any required kernel arguments. For example:
    :linux firstboot

Important

For more information on the kernel parameters, refer to Section 5.2, “RHEV Hypervisor Kernel Parameters and Automated Installation”.

Note

To upgrade an existing hypervisor, you must boot the kernel with the upgrade parameter. This will automatically upgrade and reboot the system, rather than taking you to the interactive configuration menu. For more information, refer to Section 5.2.6, “Other parameters”.

Note

You can append the kernel boot arguments to the PXE configuration file (/pxelinux.cfg/default). This can be used to run an automated setup, covered in Section 5.2, “RHEV Hypervisor Kernel Parameters and Automated Installation”, and may be more useful depending on your environment.

5.1.2. The RHEV Hypervisor Configuration Menu

Once the RHEV Hypervisor boots you will see the Hypervisor Configuration Menu.
The Hypervisor Configuration Menu is an interactive tool for installing Red Hat Enterprise Virtualization Hypervisors. The Hypervisor Configuration Menu resembles Example 5.1, “Red Hat Enterprise Virtualization Hypervisor Configuration Menu”. Throughout this chapter it will be referred to as the main menu.

Example 5.1. Red Hat Enterprise Virtualization Hypervisor Configuration Menu

Red Hat Enterprise Virtualization Hypervisor release 5.7-2.2

Hypervisor Configuration Menu

1) Configure storage partitions    6) Configure the host for RHEV
2) Configure authentication        7) View logs
3) Set the hostname                8) Install locally and reboot
4) Networking setup                9) Support Menu
5) Register Host to RHN
Choose an option to configure:

Configuration menu options

Configure storage partitions
Prepares the local storage device for installation. Refer to Section 5.1.4, “Partitioning the disk”.
Configure authentication
Sets the root password for administration and SSH access. Refer to Section 5.1.5, “Set the administrator password”.
Set the hostname
Sets the hostname manually instead of using DHCP. Refer to Section 5.1.6, “Set the hostname”.
Networking setup
Configures network settings and services including: DHCP, IPv4, NTP and DNS. Refer to Section 5.1.7, “Set up Networking”.
Register Host to RHN
Registers the host to the Red Hat Network (RHN). Refer to Section 5.1.8, “Register to RHN”.
Configure the host for RHEV
Enter IP address of the RHEV Manager server. Refer to Section 5.1.9, “Configure the host for RHEV”.
View logs
View log files to debug the installation procedure. Refer to Section 5.1.10, “View logs”.
Install locally and reboot
Install the Hypervisor to the storage device and reboot. Refer to Section 5.1.11, “Installation”.
Support Menu
Contains options for continuing stateless boot and exiting to a shell prompt. This menu is not supported except when working with Red Hat support.

5.1.3. Remove existing RHEV Hypervisor installations

If the Hypervisor was previously installed on the local system, remove the existing installation by selecting the Support Menu option from the main menu, then selecting Uninstall an existing Hypervisor.

Important

On systems that have a RHEV Hypervisor installed and some systems with LVM volumes, the firstboot parameter may be required to trigger the interactive installation. To use the firstboot parameter, append it to other kernel parameters. On a system with an existing RHEV Hypervisor, use the following kernel parameters to activate the interactive installation:
: linux firstboot
For more information on uninstalling previous versions of the RHEV Hypervisor refer to Appendix A, Uninstalling Red Hat Enterprise Virtualization Hypervisors from local storage devices.

5.1.4. Partitioning the disk

The partitions used by the RHEV Hypervisor are configured from this menu. For most installations the default partition layout is optimal.

Warning

Running this procedure will destroy all data on the local storage device selected with the storage_init parameter.

Procedure 5.2. The Storage Configuration menu

  1. Select the Configure storage partitions option from the main menu to open the Configure storage partitions menu.

    Example 5.2. The Configure storage partitions menu

     Configure storage partitions
    
    1) Configure
    2) Review
    3) Commit configuration
    4) Return to the Hypervisor Configuration Menu
    Choose an option:
    
  2. Select storage device

    Select Configure.
    If you have more than one storage device available, you are prompted to select the storage device on which to install the RHEV Hypervisor. If there is only one storage device, skip to Step 3.

    Example 5.3. Select storage device

    /dev/mapper/SServeRA_venh_076A0444 (sda) (69889 MB)
    Disk Identifier: storage_serial_SServeRA_venh_076A0444
    /dev/sdb (15296 MB)
    Disk Identifier: storage_serial_UFD_USB_Flash_Drive_XFHO0KCK69T746WFHAUB
    1) /dev/mapper/SServeRA_venh_076A0444  3) Abort
    2) /dev/sdb                            4) Manual selection
    
    Select the appropriate storage device from the menu. If the device you wish to use is not shown, select Manual selection. You will then be prompted to enter0 the path to the storage device. The path must be entered exactly; wildcards are not permitted.
  3. Choose partition sizes

    You are prompted to set the size of the boot, swap, root, config, logging and data partitions. The value is accepted in megabytes (MB).
    Setting a size value of -1 sets the partition to take up all remaining disk space. Note that this can only be used with the Data partition.
    Pressing Enter without any value sets the partition to the default size.

    Important

    Do not append units onto the end of the partition size values. Using MB or similar unit on the end of a value may have adverse effects. For example, use the value 5 instead of 5MB.
    1. The boot partition contains the boot loader, kernel and initramfs.
      The default and only supported size for the boot partition is 50MB. The size of the boot partition cannot be configured.
    2. The swap partition is used for swapping underused memory to the hard drive to speed up memory performance. The default size of the swap partition is calculated based on the amount of RAM and the overcommit ratio (default is 0.5). RHEV Hypervisors must have a swap partition and the swap partition cannot be disabled by setting its size to 0. The minimum size for the swap partition is 5MB.
      Red Hat Knowledgebase has an article on determining an appropriate size for the swap partition.
      Use the formula from the Red Hat Knowledgebase and add storage for the overcommit ratio (RAM multiplied by the overcommit ratio).
      Recommended swap + (RAM * overcommit) = swap partition size
      Pressing Enter during the installation sequence sets a recommended value for the swap partition.
    3. The root partition contains the core files for the Hypervisor and the LiveOS directory from the installation media. The default total size of the root file system partition is 512MB. This value includes 256MB for one root partition and an additional 256MB for a backup root partition.
      The default and only supported size for the root partition is 256MB. The size of the root partition cannot be configured.
    4. The config partition stores configuration files for the Hypervisor.
      The default and minimum size for the configuration partition is 5MB.
    5. The logging partition stores all logs for the Hypervisor.
      The logging partition is required and the recommended size is 2048MB.
    6. The data partition stores core files and kdump files and acts as temporary storage for ISO files. The default size of the data partition is the remaining available disk space (labeled as -1).
      When ISO files are uploaded using the ISO Uploader tool, they are temporarily stored on the data partition of a RHEV Hypervisor host, before being moved to the shared ISO Library. The data partition must include sufficient space to store these ISO files, which can be over 4GB in size for modern operating systems.
      The data partition must be large enough to hold core files for KVM. Core files depend on the RAM size for the virtual machine. The data partition must also be large enough to store kernel dump files, also known as kdumps. A kdump file is usually the same size as the host's total system RAM. The data partition also stores the RHEV Hypervisor ISO file for RHEV Hypervisor upgrades.
      The data partition must be at least 1.5x as large as the RAM on the host system plus an additional 512MB in size. The minimum size is 256MB.
  4. Review

    Select the Review menu option. The review menu output is similar to the screenshot below:
    The local disk will be repartitioned as follows:
    ================================================
    Physical Hard Disk:           /dev/hdb (10000 MB)
    Disk Identifier:              storage_serial_QM00002
    Boot partition size:          50 MB
    Swap partition size:          2233 MB
    Installation partition size:  256 * 2 MB
    Configuration partition size: 5 MB
    Logging partition size:       2048 MB
    Data partition size:          5152 MB
  5. Select Commit configuration to accept the partitions. The following warning displays:
  6. !!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!
    !!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!
    !!WARNING!!                                                    !!WARNING!!
    !!WARNING!!                                                    !!WARNING!!
    !!WARNING!! If you proceed, all data on your selected storage  !!WARNING!!
    !!WARNING!! device will be destroyed and your hard disk        !!WARNING!!
    !!WARNING!! will be irreversibly reconfigured.                 !!WARNING!!
    !!WARNING!!                                                    !!WARNING!!
    !!WARNING!!                                                    !!WARNING!!
    !!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!
    !!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!WARNING!!
    
    Do you wish to proceed([Y]es or [N]o)?
  7. Press Y to proceed and format the partitions.
Partitions are created and file systems are formatted. Once complete, the main menu reappears.

5.1.5. Set the administrator password

Setting a strong and easy to remember administrator password is an essential part of securing your infrastructure.
Setting an administrator password is not required if you do not need SSH remote access. SSH remote access is not supported unless you are instructed to enable it by Red Hat support. In normal usage scenarios you should not need to set a root password.

Procedure 5.3. Set the administrator password

  1. Select Configure authentication from the main configuration menu.

    Example 5.4. The Configure authentication menu

    SSH remote access is currently disabled
    
    1) Set administrator password
    2) Toggle SSH password authentication
    3) Return to the Hypervisor Configuration Menu
    Choose an option to configure:
  2. Select Set the administrator password.
  3. The password prompt displays.
    Set the system administrator's (root) password:
    New UNIX password:
    Enter the preferred password. You will be prompted to retype the password.
  4. Changing password for user root.
    passwd: all authentication tokens updated successfully.
    storing to /config :
    /etc/shadow File persisted
    This message indicates that the password has been set and saved to local storage.

Procedure 5.4. Optional: Configure SSH password authentication

SSH password authentication is required for accessing RHEV Hypervisors with the SSH protocol. By default, SSH password authentication is disabled. SSH remote access is not supported unless you are instructed to enable it by Red Hat support.
Follow these instructions only if instructed to do so by Red Hat support.
  1. Select Toggle SSH password authentication from the Configure authentication menu.
  2. Enable SSH password authentication ([Y]es or [N]o)?
    Press y to enable SSH password authentication.
  3. If successful, the following message is displayed:
    SSH remote access is currently enabled.
Select Return to the Hypervisor Configuration Menu to continue the installation.

5.1.6. Set the hostname

The Red Hat Enterprise Virtualization Manager requires a fully resolvable hostname for each RHEV Hypervisor. If you are using DHCP and DNS, a hostname can be assigned to the Hypervisor at boot time.
If the DHCP server does not set hostnames, the hostname must be specified manually (with the process below). This hostname must resolve to the assigned IP address located with a DNS lookup request.
  1. Select the Set the hostname option from the main menu.
  2. Enter the hostname in the following format: anyname.domain.example.com. Where anyname is the individual name for the host on the specified domain.
    What is this Hypervisor's hostname?
    Enter the hostname for the Hypervisor and press Enter.
  3. If the hostname setting procedure is successful, this message is displayed:
    The hostname is set.
You are returned to the Configuration menu.

Procedure 5.5. Resetting hostnames

To reset a hostname:
  1. Open the Set the hostname menu.
  2. The specify a hostname prompt appears:
    What is this Hypervisor's hostname?
    Press the Enter key without typing anything else to reset the hostname.
The hostname is now reset to use DHCP.

5.1.7. Set up Networking

Configure your system's network settings.

Important

For installation purposes only a single network can be configured for access to the RHEV Manger. Once the RHEV Hypervisor is installed, additional network interfaces can be configured with the RHEV Manager user interface.
  1. Select the Networking setup option.
  2. The network devices attached to your system are listed.
  3. Choose an interface or a configuration option:
    Select the network interface which will be used to connect to the RHEV Manager. On systems with one network interface there is only eth0 to choose from.

    Important

    Wireless and logical network devices are presently unsupported.
  4. Identifying the interface

    Hypervisors include a tool to assist in physically identifying which network interface was selected.
    Help identify eth0 by blinking the lights 
    for 10 seconds ([Y]es or [N]o)?
    Press y to identify which socket eth0 is. This is useful for systems with multiple network interfaces.
    If assistance is not required, continue by pressing n.
    If the selected network interface does not support this feature, the system will display the following message:
    Blinking is not supported by your network interface.
  5. VLANs

    A prompt will be displayed asking whether to include VLAN support:
    Include VLAN support ([Y]es, [N]o or [A]bort)?
    • If VLAN support is not required, press n.
    • To enable VLAN support, enter y.
      1. What is the VLAN ID for this device (a=abort)
        Enter the VLAN ID number when prompted.
  6. IPv4

    The IPv4 prompt appears.
    Enable IPv4 support ([S]tatic IP, [D]HCP, [N]o or [A]bort)?
    Select DHCP or Static IP depending on your networking environment.
    • Press D to enable IPv4 support with DHCP. It is required to statically map all IP addresses for each RHEV Hypervisor on your DHCP server.
    • Alternatively, assign a static IP address by pressing S. You are then prompted for the IP address, Netmask and gateway.
    Confirm your settings.
    Is this correct ([Y]es, [N]o or [A]bort)?
    If the settings are correct, press Y to continue.
  7. DNS

    Configure DNS if using static network configuration or of your DHCP server is not assigning DNS servers.
    1. Enter the IP address of the first DNS gateway or press Enter to abort.
      Enter the first DNS server (or press the Enter key to exit):
    2. Enter the IP address of the second DNS gateway or press Enter to skip.
      Enter the second DNS server (or press the Enter key to exit):
    3. Confirm your settings.
      Is this correct ([Y]es, [N]o or [A]bort)?
      Press Y if correct to continue.
  8. Network Time Protocol (NTP)

    The Network Time Protocol (NTP) keeps system clocks in time with the rest of the world. NTP is crucial for some host systems with inconsistent clocks, as inaccurate timekeeping prevents Hypervisors from migrating virtual machines.
    1. Enter an NTP server:
      Enter the IP address or hostname of an NTP server, or press Enter to abort.
    2. You will be prompted to enter additional NTP servers. Enter as many as required and press Enter on a blank line to save your selection.
  9. Select Save And Return To Menu to save the configuration.
  10. If the network is configured correctly, the output should be similar to Figure 5.1, “Working network configuration”.
    Working network configuration

    Figure 5.1. Working network configuration

You are now returned to the main menu.

5.1.8. Register to RHN

This procedure registers the host to RHN. Registering the host to RHN allows you to apply Red Hat Enterprise Linux virtualization or virtualization platform entitlements to the Hypervisor host. If you are not using these entitlements, you do not need to register the host to RHN.
  • Select Register Host to RHN from the main menu.
    • To register to the public RHN system, select 1) Register to public RHN
      1. Enter RHN account username:
        Enter your RHN account username.
      2. Enter password for RHN account:
        Enter your RHN account password.
      3. Enter profile name for this system (optional):
        Enter a profile name for this system. This parameter is optional. If a blank profile name is provided, the system hostname will be used by default.
      4. Enter HTTP proxy in host:port format, e.g. proxy.example.com:3128 (leave blank if not required):
        Enter your HTTP proxy address. This parameter is optional. If a blank proxy address is provided, no proxy server will be used.
    • To register to an RHN Satellite, select 2) Register to RHN Satellite
      1. Enter URL of RHN Satellite (ie. https://satellite.example.com)
        Enter your the URL of your RHN Satellite.
      2. Enter location of satellite cert (https://satellite.example.com/pub/RHN-ORG-TRUSTED-SSL-CERT):
        Enter the URL of your satellite certificate, or press enter to select the default URL shown.
      3. Enter RHN account username:
        Enter your RHN account username.
      4. Enter password for RHN account:
        Enter your RHN account password.
      5. Enter profile name for this system (optional):
        Enter a profile name for this system. This parameter is optional. If a blank profile name is provided, the system hostname will be used by default.
      6. Enter HTTP proxy in host:port format, e.g. proxy.example.com:3128 (leave blank if not required):
        Enter your HTTP proxy address. This parameter is optional. If a blank proxy address is provided, no proxy server will be used.

5.1.9. Configure the host for RHEV

This procedure connects the RHEV Hypervisor to a RHEV Manager.
When specifying the address of the RHEV Manager, use the following format for IP addresses or hostnames.

Important

Ensure that you have configured the local storage and networking correctly before proceeding. If something is not correct change it now because you cannot change it after this point. To change a setting after this point you must repeat the procedure from the beginning.
Ensure that the IP address or host name specified is fully resolvable.

Table 5.1. Red Hat Enterprise Virtualization Manager addressing examples

Type Format Example
IP address IP[:port] 127.0.0.1:443
Host Name host.name.domain[:port] www.redhat.com:443
  1. Select Configure the host for RHEV from the main menu.
  2. Connecting RHEV Manager

    Once the Configure the host for RHEV menu item is selected the following output displays:
    Enter the RHEV Manager's hostname or IP address.
    Optionally: append a port after the hostname or IP address
    For example, 10.0.0.1:443 or rhev.example.com:443
    1. Enter the RHEV Manager's hostname or IP address in the specified format.
      If the RHEV Manager port is skipped the setup will use the default port, port 443. However, if a port has previously been assigned, the RHEV Hypervisor uses that port as the default instead. For example, if port 12345 was specified during a previous installation then the default port for future installations is port 12345.

      Note

      RHEV Managers use port 443 by default. If you configured another port when installing the RHEV Manager, use it instead. Once another port is specified for the RHEV Hypervisor, that port will be recorded as the new default port.
    2. If the RHEV Manager is successfully located, the following output is displayed:
      The RHEV Manager's address is set.
      The RHEV Manager's port is set.
  3. NetConsole access

    Enter the NetConsole manager's hostname or IP address.
    Optionally: append a port after the hostname or IP address
    For example, 10.0.0.1:25285 or rhev.redhat.com:25285
    If you are unsure, use the RHEV Manager's address, without a port.
    1. Enter the NetConsole server and port. In most cases, this server is the RHEV Manager. The NetConsole port is the NetConsole port number specified during the installation of the RHEV Manager.
    2. If the NetConsole server is successfully found, the following output is displayed before being returned to the configuration menu:
      The NetConsole manager address is set.
      The NetConsole manager port is set.

5.1.10. View logs

Various log files can viewed to assist in debugging the installation procedure. Once a file is selected by its menu option number, it is displayed in full on the screen. You can scroll through the file contents by pressing Shift+Page Up to scroll up and Shift+Page Down to scroll down. The log files available are:
/var/log/ovirt.log
Contains detailed information about the hypervisor installation and configuration procedure.
/var/log/messages
Contains general messages from the system and core services.
/var/log/secure
Contains authentication and security messages, including failed login attempts.

5.1.11. Installation

Once all the previous steps are complete, the RHEV Hypervisor can be installed to the local storage device.
Storage, networking and the RHEV Manager must be configured before continuing.
  1. Select Install locally and reboot from the Hypervisor Configuration Menu.
  2. This option exits the menu and requires a reboot.
    Ensure local storage and networking is configured
    correctly on the previous menu before continuing.
    
    The current kernel boot parameters are
    crashkernel=128M@16M elevator=deadline processor.max_cstate=1
    
    Enter additional kernel boot parameters (press Enter to use the present arguments):
    Press Enter to continue.
  3. Do you wish to proceed ([Y]es or [N]o)?
    Press Y to continue.
The final installation sequence begins. Once the installation is complete the RHEV Hypervisor will reboot.
After rebooting, the RHEV Hypervisor is installed and integrated with Red Hat Enterprise Virtualization Manager for Servers. You should have a new host available in the Red Hat Enterprise Virtualization Manager interface.

Note

You may need to remove the boot media and change the boot device order to prevent the installation sequence restarting after the system reboots.

5.2. RHEV Hypervisor Kernel Parameters and Automated Installation

This section covers the kernel command line parameters for RHEV Hypervisors. These parameters can be used to automate installation. The parameters are described in detail and an example is provided for an automated installation.
This installation method is an alternative to the interactive installation covered by Section 5.1.2, “The RHEV Hypervisor Configuration Menu”. Using the method covered in this chapter with a PXE server can, with some configuration, deploy multiple RHEV Hypervisors without manually accessing the systems.
It is important to understand how the parameters work and what effects they have before attempting automated deployments. These parameters can delete data from existing systems if the system is configured to automatically boot with PXE.

5.2.1. How the kernel arguments work

Below is a description of the RHEV Hypervisor start up sequence. This may be useful for debugging issues with automated installation.
  1. The ovirt-early script sets storage, network and management parameters in the /etc/default/ovirt file. These parameters are determined from the kernel arguments passed to the Hypervisor during the boot sequence.
  2. The /etc/init.d/ovirt-firstboot script executes special RHEV scripts and start up procedures.
  3. An automated installation begins if all the required parameters are set.

5.2.2. Required parameters

At a minimum, the following parameters are required for an automated installation:
  1. storage_init to initialize a local storage device.
  2. BOOTIF to specify the network interface which the Hypervisor uses to connect to the Manager. When using PXE boot, BOOTIF may be automatically supplied by pxelinux.
  3. management_server to specify the Manager server.

5.2.3. Storage parameters

The following parameters configure local storage devices for installing a RHEV Hypervisor.

5.2.3.1. storage_init

The storage_init parameter is required for an automated installation.
The storage_init parameter initializes a local storage device. Presently, RHEV Hypervisors use one storage device for local installation. There are four methods for defining which disk to initialize and install on.
  1. For USB storage devices, use the usb parameter to select disk type, as reported by udev ID_BUS. Optionally, append a serial number to explicitly define which device to use. For example:
    storage_init=usb:Generic_STORAGE_DEVICE_0000145418-0:0
  2. For SCSI hard drives, use storage_init=scsi:serialnumber. Optionally, append a serial number to explicitly define which device to use. For example:
    storage_init=scsi:SCSI_HITACHI_HTS5416_SBADDFF
  3. For CCISS devices, use storage_init=cciss:serialnumber. Optionally, append a serial number to explicitly define which device to use. For example:
    storage_init=cciss:3600508b100104a3953545233304c0003
  4. Alternatively, the storage device can be specified by using the Linux device name as the storage_init parameter. Using device names in the format /dev/disk/by-id is not supported. storage_init must use the format /dev/mapper/$DISK or /dev/$DISK. In this example the /dev/sda device is specified:
    storage_init=/dev/sda
    Note that these device names are not guaranteed to refer to the same device every time the system boots.

Note

Serial numbers can be found with this script. The script lists serial numbers for all disks attached to the system.
for d in /dev/sd?; do eval $(udevinfo -q env -n $d); echo $d $ID_SERIAL;done

5.2.3.2. storage_vol

The storage_vol parameter is used to partition the storage device set by the storage_init parameter.
After storage_vol= there are six values separated by colons.
Values can be left undefined. If a value is not defined the installation will use the default value. The first and third values represent the boot and root partitions respectively, which have fixed sizes. These values cannot be set and should be left undefined.
All values are in megabytes (MB). Do not append units onto the end of the values.
Setting a size value of -1 sets the partition to take up all remaining disk space. Note that this can only be used with the Data partition.
The following is the standard format of the storage_vol parameter with each element described in the list below.

Example 5.5. Format of the storage_vol parameter

storage_vol=:SWAP::CONFIG:LOGGING:DATA
SWAP
The swap partition is used for swapping underused memory to the hard drive to speed up memory performance. The default size of the swap partition is calculated from amount of RAM and overcommit ratio (default is 0.5). RHEV Hypervisors must have a swap partition and the swap partition cannot be disabled by setting its size to 0. The minimum size for the swap partition is 5MB.
Red Hat Knowledgebase has an article on determining the size of the swap partition.
Use the formula from the Red Hat Knowledgebase and add storage for the overcommit ratio (RAM multiplied by the overcommit ratio).
Recommended swap + (RAM * overcommit) = swap partition size
Leaving the value empty sets a recommended value for the swap partition.
CONFIG
The config partition stores configuration files for the Hypervisor. The default and minimum size for the configuration partition is 5MB.
LOGGING
The logging partition stores all logs for the Hypervisor. The logging partition is required and the recommended size is 2048MB.
DATA
The data partition must be large enough to hold core files for KVM. Core files depend on the RAM size for the virtual machine. The data partition must also be large enough to store kernel dump files, also known as kdumps. A kdump file is usually the same size the host's system RAM. The data partition also stores the RHEV Hypervisor ISO file for RHEV Hypervisor upgrades.
The data partition should be at least 1.5x as large as the RAM on the host system plus an additional 512MB in size. The minimum size is 256MB.
The default size for the data partition is the remaining available disk space (labeled as -1).

Example 5.6. Example: Using the storage_vol parameter to partition default sizes

storage_vol=:::::

Example 5.7. Example: Using the storage_vol parameter to partition certain sizes sizes

We want to set the storage_vol parameter such that the following are true:
  • The Boot partition is defined as the fixed size of 50MB.
  • The Swap partition is defined as 4000MB.
  • The Root partition is defined as the fixed size of 512MB.
  • The Config partition is defined as 5MB.
  • The Logging partition is defined as 2048MB.
  • The Data partition is defined to take up all remaining disk space.
Therefore, the storage_vol parameter will look like this:
storage_vol=:4000::5:2048:-1

5.2.4. Networking and RHEV Manager parameters

Several networking options are available. The following information must be appended for the Hypervisor to automatically install:
  • Setting the IP address or DHCP.
  • Setting the hostname if the hostname is not resolved with DHCP.
  • The interface the RHEV Manager network is attached to.

Networking and RHEV Manager parameters

BOOTIF
Required for automated installations. BOOTIF specifies the network interface that the Hypervisor uses to connect to the RHEV Manager. It takes arguments in one of three forms:
link
Indicates to use the first interface (as enumerated by the kernel) with an active link. This is useful for systems with multiple network interface controllers but only one plugged in.
eth# (where # is the number of the NIC)
Indicates to use the NIC as determined by the kernel driver initialization order. To determine the number boot into RHEV and select Shell from the Hypervisor Configuration Menu. Use ifconfig | grep eth* to list the network interfaces attached to the system. There is no guarantee that on the next reboot the network interface controller will have the same eth# mapping.
BOOTIF=eth0
<MAC>
Indicates to use the MAC address explicitly defined inside the brackets.

Important

When using PXE to boot Hypervisors for installation using the IPAPPEND 2 directive causes BOOTIF=<MAC> to be automatically appended to the kernel arguments. If you use the IPAPPEND 2 directive you do not need to use the BOOTIF parameter.
ip
Sets the IP address for the network interface controller defined by the BOOTIF parameter. The ip parameter accepts either an IP address (in the form 0.0.0.0) or the value dhcp (for DHCP).
ip=192.168.1.1
ip=dhcp
netmask
The netmask parameter sets the subnet mask for the IP address defined with the ip parameter.
netmask=255.255.255.0
gateway
The gateway parameter sets the Internet gateway.
gateway=192.168.1.246
dns
The dns parameter sets the addresses of one or more DNS servers. Each DNS server must be separated by a colon.
dns=192.168.1.243:192.168.1.244
hostname
The hostname parameter sets the hostname. The hostname must be a fully-qualified and resolvable domain name.
hostname=rhev1.example.com
ntp
The ntp parameter sets the addresses of one or more Network Time Protocol servers. Each NTP server must be separated by a colon.
ntp=192.168.2.253:192.168.2.254
vlan
The vlan parameter sets the VLAN identifier for the network connected to the RHEV Manager. Set this parameter if you are using VLANs.
vlan=VLAN-ID

5.2.5. Red Hat Network (RHN) parameters

These parameters are used to automatically register the hypervisor host with the Red Hat Network (RHN). At a minimum, either the rhn_activationkey or both the rhn_username and rhn_password parameters must be provided. If you are registering to a satellite server, you must also provide a value for the rhn_url parameter.
rhn_username
The rhn_username parameter sets the username used to connect to RHN.
rhn_username=testuser
rhn_password
The rhn_password parameter sets the password used to connect to RHN.
rhn_password=testpassword
rhn_activationkey
The rhn_activationkey parameter sets the activation key used to connect to RHN. Activation keys are used to register systems, entitle them to an RHN service level, and subscribe them to specific channels and system groups, all in one action. If both rhn_activationkey and rhn_username are provided, the rhn_activationkey value will be used.
rhn_activationkey=7202f3b7d218cf59b764f9f6e9fa281b
rhn_url
The rhn_url parameter sets the URL of the satellite server used to register the host.
rhn_url=https://your-satellite.example.com
rhn_ca_cert
The rhn_ca_cert parameter sets the URL of the CA certificate used to connect to the satellite server. If it is not provided, the default value is rhn_url/pub/RHN-ORG-TRUSTED-SSL-CERT
rhn_ca_cert=https://your-satellite.example.com/pub/RHN-ORG-TRUSTED-SSL-CERT
rhn_profile
The rhn_profile parameter sets the name of the profile to be registered with RHN for this host. The default value is the system hostname.
rhn_profile=testhost

5.2.6. Other parameters

management_server
The management_server parameter is required for an automated installation.
The management_server parameter sets the address of the RHEV Manager. The management_server parameter takes an IP address or fully qualified domain name and, optionally, a port (the default port is 443). It is required for an automated installation.
management_server=rhev.example.com:443
netconsole_server
The netconsole_server parameter sets the address of the RHEV Manager. The netconsole_server parameter takes an IP address or fully qualified domain name and, optionally, a port (the default port is 25285).
netconsole_server=rhev.example.com:25285
mem_overcommit
The mem_overcommit parameter specifies the multiplier to use for adding extra swap to support memory overcommit. The default overcommit value is 0.5.
mem_overcommit=0.7
firstboot
The firstboot parameter starts the interactive configuration menu. On systems that have a RHEV Hypervisor installed and some systems with LVM volumes, the firstboot parameter may be required to trigger the interactive installation. Refer to Section 5.1.2, “The RHEV Hypervisor Configuration Menu” for installation from the configuration menu.
upgrade
The upgrade parameter will upgrade the existing hypervisor image to the version provided by the boot media. The hypervisor will be automatically upgraded and rebooted once complete. If a hypervisor image is not yet installed, the image will be installed to the device selected with the storage_init parameter. When performing an upgrade, the previous boot entry is saved as BACKUP in grub.conf. If the reboot following the upgrade procedure fails, the BACKUP boot entry will be automatically selected as the new default.
local_boot
The local_boot parameter is an alias for the upgrade parameter.
nocheck
The nocheck parameter will skip the MD5 check of the installation ISO, which can be time consuming if the media is remote or slow.
qemu_pxe
The qemu_pxe parameter is used to select which network bootloader is used in virtual machines. Possible values are gpxe and etherboot. For compatibility with RHEV Hypervisor 5.4-2.1, the default value is etherboot.
qemu_pxe=gpxe

5.2.7. Example: Automated RHEV Hypervisor Installation

This example uses the kernel command line parameters for an automated RHEV Hypervisor installation. Ensure you customize the example for your environment as not all installations are the same.

Warning

This example may not work accurately on all systems. Review the descriptions of the parameters above and modify as required for your systems.
The following is a typical example for installing a RHEV Hypervisor with the kernel command line parameters
In this example, the RHEV Manager is located at the hostname: rhevm.example.com.
:linux storage_init=/dev/sda storage_vol=::::: local_boot BOOTIF=eth0 management_server=rhevm.example.com netconsole=rhevm.example.com

Note

The kernel parameters can be automatically appended to virtual machines booting over a network with PXE. Automatically installing from PXE is not covered by this guide.

Chapter 6. Maintaining and Administrating Red Hat Enterprise Virtualization Hypervisors

This chapter covers administration and maintenance tasks for Red Hat Enterprise Virtualization Hypervisors.

6.1. Enabling kdumps

RHEV Hypervisor hosts generate a kernel dump (a kdump file) in the event of a system failure. These kdump files are essential for debugging and support.
All kdump files are placed on the /data partition of the storage on the RHEV Hypervisor host. When installing a RHEV Hypervisor host, ensure there is sufficient persistent storage space for kdump files. The size of the kdump file is proportional to the amount of RAM installed on the host. For a system with 32GB of RAM, a Hypervisor requires a /data partition of at least 32GB plus additional space for RHEV specific data (approximately 512MB).
The kdump file can be transferred off the RHEV Hypervisor host once the host is rebooted and is running properly. To transfer the kdump file, use the following procedure.

Note

There is an Open Source SSH client for Windows systems called PuTTY. Download the client from the PuTTY homepage: http://www.chiark.greenend.org.uk/~sgtatham/putty/. Although many of the steps remain the same, using PuTTY to transfer files is not covered in this guide.
  1. Access a Linux, UNIX, or similar operating system with SSH and the scp command installed. The scp command has the following format:
    scp source-user@source_IP_OR_hostname:/path/to/files destination-user@destination_IP_OR_hostname:/path/to/destination
  2. This example copies all kdump files to the /tmp/RHEVkdumps directory on the local system. Use the administrator (root) username and password configured during RHEV Hypervisor installation. Replace the example hostname, rhevh.example.com, with a valid IP address or hostname for the RHEV Hypervisor.
    $ scp -r root@rhevh.example.com:'/data/core/*' /tmp/RHEVkdumps
    
  3. Once this procedure completes you can delete all the original kdump files on the Hypervisor. Log into the Hypervisor as root and execute the following.
    # rm -rf /data/core/*
    
The kdump file is now removed and available on the other host for support and debugging.

6.2. Setting the root password

By default, the root login and password are disabled on RHEV Hypervisors for security reasons. During unattended installs, there is no way to override this and set the root password at install time. In some circumstances, however, having root access to the RHEV Hypervisor host is necessary for debugging and support. To work around this, administrators must first gain temporary root access to the RHEV Hypervisor host, then set the password.
  1. Gaining temporary root access

    • Method 1: SSH using a key pair. This procedure is run from the RHEV Manager server.
      1. Download the latest versions of putty.exe and puttygen.exe to your RHEV Manager server from http://www.chiark.greenend.org.uk/~sgtatham/putty/download.html
      2. Use puttygen to generate a new private key.
        1. Run puttygen.exe. From the top menu, select conversion -> Import key
        2. Select C:\Program Files\RedHat\RHEVManager\Service\ca\keys\rhevm.pem
        3. Enter the passphrase for this key: mypass
        4. Click save private key
      3. Use the private key to connect to the RHEV Hypervisor host.
        1. Run putty.exe. Enter the IP address of the RHEV Hypervisor host you wish to connect to.
        2. From the left panel select SSH -> Auth
        3. Browse to the private key you just generated, click open
        4. Enter the username root and use the password mypass. Open the SSH connection. You should now have a root shell open on the RHEV Hypervisor host.
    • Method 2: Boot into rescue mode. This procedure requires physical access to the RHEV Hypervisor host, and is run from its physical console.
      1. Reboot the machine and wait for the GRUB prompt. Hit enter to interrupt normal booting.
      2. Modify the kernel line to boot into rescue mode. To do this, append the string rescue to the kernel line, separated from the rest of the line with a space character. Once the system boots, you will be dropped to a a single user root shell. The system must be rebooted into multi-user mode after following the instructions below.
  2. Setting the root password

    Once you have successfully gained root access, you can set the root password for future use.
    1. Run the command setup
    2. Select option #1 Configure authentication
    3. Select option #2 Toggle SSH password authentication
    4. Enable SSH password authentication
    5. Select option #1 Set administrator password
    6. Enter your new root password
    7. Select option #3 Return to the Hypervisor Configuration Menu
    8. Select option #7 Exit Hypervisor Configuration Menu

Chapter 7. Upgrading Red Hat Enterprise Virtualization Hypervisors

Red Hat Enterprise Virtualization (RHEV) Hypervisors can be updated to get the latest features, bug fixes and security patches.

7.1. Upgrading a RHEV Hypervisor with a RHEV Manager

You can upgrade and reinstall a Red Hat Enterprise Virtualization Hypervisor host from an ISO stored on the RHEV Manager. Upgrading and reinstalling means that you are stopping and restarting the host. Virtual machines are automatically migrated to a different host, however it is recommended that the upgrade is performed at a time when usage of the system is at its lowest. Ensure that the cluster contains more than one host before performing an upgrade.
It is recommended that administrators update Red Hat Enterprise Virtualization Hypervisors regularly. Important bug fixes and security updates are included in updates. Hypervisors which are not up to date may be a security risk.

Warning

Upgrading hosts involves shutting down, deactivating virtual machines and restarting the physical server. If any virtual machines are running on the Hypervisor, all data and configuration details may be destroyed if they are not shut down correctly. Upgrading Hypervisors must be carefully planned and executed with care and consideration.

7.1.1. Prerequisites

Before upgrading a Hypervisor:
  1. Download the latest RHEV Hypervisor RPM package from Red Hat Network.
    Or, download the latest RHEV Hypervisor ISO image from Red Hat Network
  2. Copy the ISO image file to the C:\Program Files\RedHat\RHEVManager\Service\RHEV-H Installer folder on the Red Hat Enterprise Virtualization Manager server.
  3. Change the file permissions so that NETWORK SERVICE has full control of the iso file.
    1. Use Windows Explorer to navigate to C:\Program Files\RedHat\RHEVManager\Service\.
    2. Right-click on the updated file, for example, rhev-hypervisor.iso, and select the Properties option. The Properties dialog displays.
    3. Click the Security tab, and select NETWORK SERVICE from the Group or user names panel.
    4. In the Permissions panel, click the Allow Full Control check box.
    5. Click OK.

Important

Ensure that the cluster contains more than one host before performing an upgrade. Do not attempt to re-install or upgrade all hosts at the same time, as one host must remain available to perform Storage Pool Manager (SPM) tasks.

7.1.2. Upgrading a RHEV Hypervisor

  1. Click the Hosts tab.
    A list of hosts displays. Select the host that you intend to upgrade.
    If the host is not displayed, perform a search.
  2. On the Details pane, click the General Tab.
    General Pane with Upgrade Button

    Figure 7.1. General Pane with Upgrade Button

    An Alert Message indicates that a new version of the RHEV Hypervisor is available. The Upgrade link is disabled if the host is has a status of Up. A tooltip directs you to switch to maintenance mode to enable upgrade.
    Upgrade Button

    Figure 7.2. Upgrade Button

  3. Select the host and click the Maintenance button. This will cause any virtual machines running on the host to be migrated to other hosts. If the host is the SPM, this function will be moved to another host. The status of the host changes as it enters maintenance mode. When the host status is Maintenance, the message in the general tab changes, providing you with a selectable link to re-install or upgrade the host.
  4. Click the link. The Install Host dialog displays.
    Install Host Dialog

    Figure 7.3. Install Host Dialog

  5. Select the appropriate file from the list of available ISOs to upgrade the host. This is usually rhev-hypervisor.iso.
  6. Click OK to upgrade and re-install the host. The dialog closes, the details of the host are updated in the Hosts tab, and the status changes appropriately.
    The host status transitions through the following stages: Installing, Reboot, Non Responsive, and Up. These are all normal, and may take some time.
  7. Once successfully upgraded, the host displays a status of Up. Any virtual machines that were migrated off the host, may be migrated back onto the upgraded host.

7.2. Upgrading a RHEV Hypervisor with local media

To upgrade an existing hypervisor from a local CD-ROM or USB device, insert the installation media and ensure the system boots from it. Boot the kernel with the upgrade parameter. This will automatically upgrade and reboot the system, rather than taking you to the interactive configuration menu. For more information, refer to Section 5.2.6, “Other parameters”.
ISOLINUX 3.11 2005-09-02  Copyright (C) 1994-2005 H. Peter Anvin
This is the Red Hat Enterprise Virtualization Hypervisor 5.7-2.2 (4.2)
- to boot press the <ENTER> key
or type linux and press the <ENTER> key.
- Extra kernel command line parameters can be passed following the 
linux keyword.
- Boot parameters summary:
BOOTIF=link|eth*|<MAC>(pxelinux format)
storage_init=[usb|scsi|cciss[:serial#]]|/dev/...
storage_vol=:SWAP_MB::CONFIG_MB:LOGGING_MB:DATA_MB
mem_overcommit=<overcommit_ratio>
upgrade
ip=<client-ip> netmask=<netmask> gateway=<gw-ip>
vlan=<VLAN-id>
dns=server[:server]
ntp=server[:server]
hostname=fqdn
management_server=RHEVM-host[:port]
netconsole_server=RHEVM-host[:port]
boot: linux upgrade

Figure 7.4. Booting with the upgrade parameter

7.3. Re-installing RHEV Hypervisors with the RHEV Manager

Re-installing RHEV Hypervisors is the same procedure as upgrading, refer to Section 7.1.2, “Upgrading a RHEV Hypervisor” for details on upgrading.
To reset settings, remove the previous Hypervisor first with the Uninstall an existing Hypervisor menu item and then re-install the Hypervisor.

Chapter 8. Security topics

The Red Hat Enterprise Virtualization Hypervisor has various security features enabled. Security-Enhanced Linux (SELinux) and the iptables firewall are fully configured and enabled by default.
Administrators can receive the latest security advisories from the Red Hat Enterprise Virtualization watch list. Subscribe to the Red Hat Enterprise Virtualization watch list to receive new security advisories for RHEV products by email. Subscribe by completing this form: http://www.redhat.com/mailman/listinfo/rhev-watch-list/.
RHEV uses various network ports for management and other virtualization features. These ports must be open for Red Hat Enterprise Linux to function as a host with Red Hat Enterprise Virtualization. The list below covers ports and their usage by Red Hat Enterprise Virtualization:
  • ICMP requests must be accepted. ICMP packets are used for network testing by the Red Hat Enterprise Virtualization Manager.
  • Port 22 should be open for SSH access and the initial installation.
  • Ports 80 or 443 (depending on the security settings on the Red Hat Enterprise Virtualization Manager) are used by the vdsm-reg service to communicate information about the host.
  • Ports 5634 to 6166 are used for guest virtual machine console access.
  • Ports 49152 to 49216 are used for migrations. Migration may use any port in this range depending on the number of concurrent migrations occurring.
  • Port 54321 is used by default, by VDSM for management, storage and inter-host communication. This port can be modified.

Appendix A. Uninstalling Red Hat Enterprise Virtualization Hypervisors from local storage devices

This appendix covers uninstalling the Red Hat Enterprise Virtualization (RHEV) Hypervisors from local storage devices.

Procedure A.1. Removing RHEV Hypervisors

  1. Boot the RHEV Hypervisor. Refer to Chapter 4, Preparing Red Hat Enterprise Virtualization Hypervisor installation media for details on booting Red Hat Enterprise Virtualization Hypervisors.
    The firstboot kernel parameter is required for uninstalling hypervisors.
    : linux firstboot
  2. Select Support Menu from the Hypervisor Configuration Menu. From the Support Menu select Uninstall an existing Hypervisor.
  3. The following message will appear:
    !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
    !!                                                                           !!
    !! WARNING !! WARNING !! WARNING !! WARNING !! WARNING !! WARNING !! WARNING !!
    !!                                                                           !!
    !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
    
       If you proceed, you will DESTROY any existing data related to the RHEV
       installation on this machine.
    
    Do you wish to continue and uninstall this Hypervisor (Y/N)?
    Press Y to confirm and remove the Hypervisor.
The RHEV Hypervisor is now removed.

Appendix B. Revision History

Revision History
Revision 5.9-6.4002013-10-31Rüdiger Landmann
Rebuild with publican 4.0.0
Revision 5.9-6Sat Jan 5 2013Laura Bailey
Publishing for Red Hat Enterprise Linux 5.9 GA.
Revision 5.9-4Wed Sep 12 2012Laura Bailey
Building for RHEL 5.9.
Revision 5.9-3Friday June 01 2012Laura Bailey
Minor rearrangement of sections to ensure that parameter types are clearly visible from table of contents.
Minor, transparent changes to XML to improve underlying code consistency (BZ#826827).
Revision 5.9-1Thursday May 31 2012Laura Bailey
Ensured that the terms used to describe virtual machines and guest operating systems were used consistently throughout the guide (BZ#813618).
Revision 5.8-9Tuesday February 21 2012Laura Bailey
Updating for Red Hat Enterprise Linux 5.8
Revision 4-0Thursday June 30 2011Stephen Gordon
Updated instructions for boot media creation. Tools are now included in the Hypervisor ISO itself, BZ#690254.
Revision 3-0Tuesday July 6 2010David Jorm
Documented required parameters for automated installation, BZ#611440
Revision 2-0Thursday June 24 2010David Jorm
Fixed rhn_url parameter example, BZ#607274
Updated boot menu text to show release build banner, BZ#594699
Revision 1-0Thursday June 10 2010David Jorm
RHEV 2.2
General Availability

Legal Notice

Copyright © 2010 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, OpenShift, 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.