Managing, monitoring, and updating the kernel

Red Hat Enterprise Linux 9

A guide to managing the Linux kernel on Red Hat Enterprise Linux 9

Red Hat Customer Content Services

Abstract

As a system administrator, you can configure the Linux kernel to optimize the operating system. Changes to the Linux kernel can improve system performance, security, and stability, as well as your ability to audit the system and troubleshoot problems.

Making open source more inclusive

Red Hat is committed to replacing problematic language in our code, documentation, and web properties. We are beginning with these four terms: master, slave, blacklist, and whitelist. Because of the enormity of this endeavor, these changes will be implemented gradually over several upcoming releases. For more details, see our CTO Chris Wright’s message.

Providing feedback on Red Hat documentation

We appreciate your feedback on our documentation. Let us know how we can improve it.

Submitting feedback through Jira (account required)

  1. Log in to the Jira website.
  2. Click Create in the top navigation bar
  3. Enter a descriptive title in the Summary field.
  4. Enter your suggestion for improvement in the Description field. Include links to the relevant parts of the documentation.
  5. Click Create at the bottom of the dialogue.

Chapter 1. The Linux kernel

Learn about the Linux kernel and the Linux kernel RPM package provided and maintained by Red Hat (Red Hat kernel). Keep the Red Hat kernel updated, which ensures the operating system has all the latest bug fixes, performance enhancements, and patches, and is compatible with new hardware.

1.1. What the kernel is

The kernel is a core part of a Linux operating system that manages the system resources and provides interface between hardware and software applications.

The Red Hat kernel is a custom-built kernel based on the upstream Linux mainline kernel that Red Hat engineers further develop and harden with a focus on stability and compatibility with the latest technologies and hardware.

Before Red Hat releases a new kernel version, the kernel needs to pass a set of rigorous quality assurance tests.

The Red Hat kernels are packaged in the RPM format so that they are easily upgraded and verified by the DNF package manager.

Warning

Kernels that have not been compiled by Red Hat are not supported by Red Hat.

1.2. RPM packages

An RPM package consists of an archive of files and metadata used to install and erase these files. Specifically, the RPM package contains the following parts:

GPG signature
The GPG signature is used to verify the integrity of the package.
Header (package metadata)
The RPM package manager uses this metadata to determine package dependencies, where to install files, and other information.
Payload
The payload is a cpio archive that contains files to install to the system.

There are two types of RPM packages. Both types share the file format and tooling, but have different contents and serve different purposes:

Source RPM (SRPM)
An SRPM contains source code and a SPEC file, which describes how to build the source code into a binary RPM. Optionally, the SRPM can contain patches to source code.
Binary RPM
A binary RPM contains the binaries built from the sources and patches.

1.3. The Linux kernel RPM package overview

The kernel RPM is a meta package that does not contain any files, but rather ensures that the following required sub-packages are properly installed:

kernel-core
Contains the binary image of the Linux kernel (vmlinuz).
kernel-modules-core
Contains the basic kernel modules to ensure core functionality. This includes the modules essential for the proper functioning of the most commonly used hardware.
kernel-modules
Contains the remaining kernel modules that are not present in kernel-core.

The kernel-core and kernel-modules-core sub-packages together can be used in virtualized and cloud environments to provide a RHEL 9 kernel with a quick boot time and a small disk size footprint. kernel-modules sub-package is usually unnecessary for such deployments.

Optional kernel packages are for example:

kernel-modules-extra
Contains kernel modules for rare hardware and modules which loading is disabled by default.
kernel-debug
Contains a kernel with numerous debugging options enabled for kernel diagnosis, at the expense of reduced performance.
kernel-tools
Contains tools for manipulating the Linux kernel and supporting documentation.
kernel-devel
Contains the kernel headers and makefiles sufficient to build modules against the kernel package.
kernel-abi-stablelists
Contains information pertaining to the RHEL kernel ABI, including a list of kernel symbols that are needed by external Linux kernel modules and a dnf plug-in to aid enforcement.
kernel-headers
Includes the C header files that specify the interface between the Linux kernel and user-space libraries and programs. The header files define structures and constants that are needed for building most standard programs.
kernel-uki-virt

Contains the Unified Kernel Image (UKI) of the RHEL kernel.

UKI combines the Linux kernel, initramfs, and the kernel command line into a single signed binary which can be booted directly from the UEFI firmware.

kernel-uki-virt contains the required kernel modules to run in virtualized and cloud environments and can be used instead of the kernel-core sub-package.

Important

kernel-uki-virt is provided as Technology Preview in RHEL 9.2.

1.4. Displaying contents of a kernel package

To determine if a kernel package provides a specific file, such as a module, you can display the file list of the package for your architecture by querying the repository. It is not necessary to download or install the package to display the file list.

Use the dnf utility to query the file list, for example, of the kernel-core, kernel-modules-core, or kernel-modules package. Note that the kernel package is a meta package that does not contain any files.

Procedure

  1. List the available versions of a package:

    $ dnf repoquery <package_name>

    For example, list the available versions of the kernel-core package:

    $ dnf repoquery kernel-core
    kernel-core-0:5.14.0-162.12.1.el9_1.x86_64
    kernel-core-0:5.14.0-162.18.1.el9_1.x86_64
    kernel-core-0:5.14.0-162.22.2.el9_1.x86_64
    kernel-core-0:5.14.0-162.23.1.el9_1.x86_64
    ...
  2. Display the list of files in a package:

    $ dnf repoquery -l <package_name>

    For example, display the list of files in the kernel-core-0:5.14.0-162.23.1.el9_1.x86_64 package.

    $ dnf repoquery -l kernel-core-0:5.14.0-162.23.1.el9_1.x86_64
    /boot/System.map-5.14.0-162.23.1.el9_1.x86_64
    /boot/config-5.14.0-162.23.1.el9_1.x86_64
    /boot/initramfs-5.14.0-162.23.1.el9_1.x86_64.img
    /boot/symvers-5.14.0-162.23.1.el9_1.x86_64.gz
    /boot/vmlinuz-5.14.0-162.23.1.el9_1.x86_64
    /lib/modules
    /lib/modules/5.14.0-162.23.1.el9_1.x86_64
    /lib/modules/5.14.0-162.23.1.el9_1.x86_64/.vmlinuz.hmac
    /lib/modules/5.14.0-162.23.1.el9_1.x86_64/System.map
    ...

1.5. Installing specific kernel versions

Install new kernels using the dnf package manager.

Procedure

  • To install a specific kernel version, enter the following command:

    # dnf install kernel-{version}

1.6. Updating the kernel

Update the kernel using the dnf package manager.

Procedure

  1. To update the kernel, enter the following command:

    # dnf update kernel

    This command updates the kernel along with all dependencies to the latest available version.

  2. Reboot your system for the changes to take effect.

Additional resources

1.7. Setting a kernel as default

Set a specific kernel as default using the grubby command-line tool and GRUB.

Procedure

  • Setting the kernel as default, using the grubby tool

    • Enter the following command to set the kernel as default using the grubby tool:

      # grubby --set-default $kernel_path

      The command uses a machine ID without the .conf suffix as an argument.

      Note

      The machine ID is located in the /boot/loader/entries/ directory.

  • Setting the kernel as default, using the id argument

    • List the boot entries using the id argument and then set an intended kernel as default:

      # grubby --info ALL | grep id
      # grubby --set-default /boot/vmlinuz-<version>.<architecture>
      Note

      To list the boot entries using the title argument, execute the # grubby --info=ALL | grep title command.

  • Setting the default kernel for only the next boot

    • Execute the following command to set the default kernel for only the next reboot using the grub2-reboot command:

      # grub2-reboot <index|title|id>
      Warning

      Set the default kernel for only the next boot with care. Installing new kernel RPM’s, self-built kernels, and manually adding the entries to the /boot/loader/entries/ directory may change the index values.

Chapter 2. The 64k page size kernel

kernel-64k is an additional, optional 64-bit ARM architecture kernel package that supports 64k pages. This additional kernel exists alongside the RHEL 9 for ARM kernel which supports 4k pages.

Optimal system performance directly relates to different memory configuration requirements. These requirements are addressed by the two variants of kernel, each suitable for different workloads. RHEL 9 on 64-bit ARM hardware thus offers two MMU page sizes:

  • 4k pages kernel for efficient memory usage in smaller environments,
  • kernel-64k for workloads with large, contiguous memory working sets.

The 4k pages kernel and kernel-64k do not differ in the user experience as the user space is the same. You can choose the variant that addresses your situation the best.

4k pages kernel

Use 4k pages for more efficient memory usage in smaller environments, such as those in Edge and lower-cost, small cloud instances. In these environments, increasing the physical system memory amounts is not practical due to space, power, and cost constraints. Also, not all 64-bit ARM architecture processors support a 64k page size.

The 4k pages kernel supports graphical installation using Anaconda, system or cloud image-based installations, as well as advanced installations using Kickstart.

kernel-64k

The 64k page size kernel is a useful option for large datasets on ARM platforms. kernel-64k is suitable for memory-intensive workloads as it has significant gains in overall system performance, namely in large database, HPC, and high network performance.

You must choose page size on 64-bit ARM architecture systems at the time of installation. You can install kernel-64k only by Kickstart by adding the kernel-64k package to the package list in the Kickstart file.

Chapter 3. Managing kernel modules

Learn about kernel modules, how to display their information, and how to perform basic administrative tasks with kernel modules.

3.1. Introduction to kernel modules

The Red Hat Enterprise Linux kernel can be extended with optional, additional pieces of functionality, called kernel modules, without having to reboot the system. On Red Hat Enterprise Linux 9, kernel modules are extra kernel code which is built into compressed <KERNEL_MODULE_NAME>.ko.xz object files.

The most common functionality enabled by kernel modules are:

  • Device driver which adds support for new hardware
  • Support for a file system such as GFS2 or NFS
  • System calls

On modern systems, kernel modules are automatically loaded when needed. However, in some cases it is necessary to load or unload modules manually.

Like the kernel itself, the modules can take parameters that customize their behavior if needed.

Tooling is provided to inspect which modules are currently running, which modules are available to load into the kernel and which parameters a module accepts. The tooling also provides a mechanism to load and unload kernel modules into the running kernel.

3.2. Kernel module dependencies

Certain kernel modules sometimes depend on one or more other kernel modules. The /lib/modules/<KERNEL_VERSION>/modules.dep file contains a complete list of kernel module dependencies for the respective kernel version.

depmod

The dependency file is generated by the depmod program, which is a part of the kmod package. Many of the utilities provided by kmod take module dependencies into account when performing operations so that manual dependency-tracking is rarely necessary.

Warning

The code of kernel modules is executed in kernel-space in the unrestricted mode. Because of this, you should be mindful of what modules you are loading.

weak-modules

In addition to depmod, Red Hat Enterprise Linux provides the weak-modules script shipped also with the kmod package. weak-modules determines which modules are kABI-compatible with installed kernels. While checking modules kernel compatibility, weak-modules processes modules symbol dependencies from higher to lower release of kernel for which they were built. This means that weak-modules processes each module independently of kernel release they were built against.

3.3. Listing installed kernel modules

The grubby --info=ALL command displays an indexed list of installed kernels on !BLS and BLS installs.

Procedure

  • List the installed kernels using the following command:

    # grubby --info=ALL | grep title

    The list of all installed kernels is displayed as follows:

    title="Red Hat Enterprise Linux (5.14.0-1.el9.x86_64) 9.0 (Plow)"
    title="Red Hat Enterprise Linux (0-rescue-0d772916a9724907a5d1350bcd39ac92) 9.0 (Plow)"

The above example displays the installed kernels list of grubby-8.40-17, from the GRUB menu.

3.4. Listing currently loaded kernel modules

View the currently loaded kernel modules.

Prerequisites

  • The kmod package is installed.

Procedure

  • To list all currently loaded kernel modules, enter:

    $ lsmod
    
    Module                  Size  Used by
    fuse                  126976  3
    uinput                 20480  1
    xt_CHECKSUM            16384  1
    ipt_MASQUERADE         16384  1
    xt_conntrack           16384  1
    ipt_REJECT             16384  1
    nft_counter            16384  16
    nf_nat_tftp            16384  0
    nf_conntrack_tftp      16384  1 nf_nat_tftp
    tun                    49152  1
    bridge                192512  0
    stp                    16384  1 bridge
    llc                    16384  2 bridge,stp
    nf_tables_set          32768  5
    nft_fib_inet           16384  1
    …​

    In the example above:

    1. The Module column provides the names of currently loaded modules.
    2. The Size column displays the amount of memory per module in kilobytes.
    3. The Used by column shows the number, and optionally the names of modules that are dependent on a particular module.

Additional resources

  • The /usr/share/doc/kmod/README file
  • The lsmod(8) manual page

3.5. Displaying information about kernel modules

Use the modinfo command to display some detailed information about the specified kernel module.

Prerequisites

  • The kmod package is installed.

Procedure

  • To display information about any kernel module, enter:

    $ modinfo <KERNEL_MODULE_NAME>

    For example:

    $ modinfo virtio_net
    
    filename:       /lib/modules/5.14.0-1.el9.x86_64/kernel/drivers/net/virtio_net.ko.xz
    license:        GPL
    description:    Virtio network driver
    rhelversion:    9.0
    srcversion:     8809CDDBE7202A1B00B9F1C
    alias:          virtio:d00000001v*
    depends:        net_failover
    retpoline:      Y
    intree:         Y
    name:           virtio_net
    vermagic:       5.14.0-1.el9.x86_64 SMP mod_unload modversions
    …​
    parm:           napi_weight:int
    parm:           csum:bool
    parm:           gso:bool
    parm:           napi_tx:bool

    You can query information about all available modules, regardless of whether they are loaded or not. The parm entries show parameters the user is able to set for the module, and what type of value they expect.

    Note

    When entering the name of a kernel module, do not append the .ko.xz extension to the end of the name. Kernel module names do not have extensions; their corresponding files do.

Additional resources

  • The modinfo(8) manual page

3.6. Loading kernel modules at system runtime

The optimal way to expand the functionality of the Linux kernel is by loading kernel modules. Use the modprobe command to find and load a kernel module into the currently running kernel.

Important

The changes described in this procedure will not persist after rebooting the system. For information about how to load kernel modules to persist across system reboots, see Loading kernel modules automatically at system boot time.

Prerequisites

  • Root permissions
  • The kmod package is installed.
  • The respective kernel module is not loaded. To ensure this is the case, list the loaded kernel modules.

Procedure

  1. Select a kernel module you want to load.

    The modules are located in the /lib/modules/$(uname -r)/kernel/<SUBSYSTEM>/ directory.

  2. Load the relevant kernel module:

    # modprobe <MODULE_NAME>
    Note

    When entering the name of a kernel module, do not append the .ko.xz extension to the end of the name. Kernel module names do not have extensions; their corresponding files do.

Verification

  • Optionally, verify the relevant module was loaded:

    $ lsmod | grep <MODULE_NAME>

    If the module was loaded correctly, this command displays the relevant kernel module. For example:

    $ lsmod | grep serio_raw
    serio_raw              16384  0

Additional resources

  • The modprobe(8) manual page

3.7. Unloading kernel modules at system runtime

At times, you find that you need to unload certain kernel modules from the running kernel. Use the modprobe command to find and unload a kernel module at system runtime from the currently loaded kernel.

Warning

Do not unload kernel modules when they are used by the running system. Doing so can lead to an unstable or non-operational system.

Important

After finishing this procedure, the kernel modules that are defined to be automatically loaded on boot, will not stay unloaded after rebooting the system. For information about how to counter this outcome, see Preventing kernel modules from being automatically loaded at system boot time.

Prerequisites

  • Root permissions
  • The kmod package is installed.

Procedure

  1. List all loaded kernel modules:

    # lsmod
  2. Select the kernel module you want to unload.

    If a kernel module has dependencies, unload those prior to unloading the kernel module. For details on identifying modules with dependencies, see Listing currently loaded kernel modules and Kernel module dependencies.

  3. Unload the relevant kernel module:

    # modprobe -r <MODULE_NAME>

    When entering the name of a kernel module, do not append the .ko.xz extension to the end of the name. Kernel module names do not have extensions; their corresponding files do.

Verification

  • Optionally, verify the relevant module was unloaded:

    $ lsmod | grep <MODULE_NAME>

    If the module was unloaded successfully, this command does not display any output.

Additional resources

  • modprobe(8) manual page

3.8. Unloading kernel modules at early stages of the boot process

In certain situations, it is necessary to unload a kernel module very early in the booting process. For example, when the kernel module contains a code, which causes the system to become unresponsive, and the user is not able to reach the stage to permanently disable the rogue kernel module. In that case it is possible to temporarily block the loading of the kernel module using a boot loader.

You can edit the relevant boot loader entry to unload the desired kernel module before the booting sequence continues.

Important

The changes described in this procedure will not persist after the next reboot. For information about how to add a kernel module to a denylist so that it will not be automatically loaded during the boot process, see Preventing kernel modules from being automatically loaded at system boot time.

Prerequisites

  • You have a loadable kernel module that you want to prevent from loading for some reason.

Procedure

  1. Boot the system into the boot loader.
  2. Use the cursor keys to highlight the relevant boot loader entry.
  3. Press the e key to edit the entry.

    Figure 3.1. Kernel boot menu

    kernel bootmenu rhel9
  4. Use the cursor keys to navigate to the line that starts with linux.
  5. Append modprobe.blacklist=module_name to the end of the line.

    Figure 3.2. Kernel boot entry

    kernel boot entry rhel9

    The serio_raw kernel module illustrates a rogue module to be unloaded early in the boot process.

  6. Press Ctrl+X to boot using the modified configuration.

Verification

  • Once the system fully boots, verify that the relevant kernel module is not loaded.

    # lsmod | grep serio_raw

Additional resources

3.9. Loading kernel modules automatically at system boot time

Configure a kernel module so that it is loaded automatically during the boot process.

Prerequisites

  • Root permissions
  • The kmod package is installed.

Procedure

  1. Select a kernel module you want to load during the boot process.

    The modules are located in the /lib/modules/$(uname -r)/kernel/<SUBSYSTEM>/ directory.

  2. Create a configuration file for the module:

    # echo <MODULE_NAME> > /etc/modules-load.d/<MODULE_NAME>.conf
    Note

    When entering the name of a kernel module, do not append the .ko.xz extension to the end of the name. Kernel module names do not have extensions; their corresponding files do.

  3. Optionally, after reboot, verify the relevant module was loaded:

    $ lsmod | grep <MODULE_NAME>

    The example command above should succeed and display the relevant kernel module.

Important

The changes described in this procedure will persist after rebooting the system.

Additional resources

  • modules-load.d(5) manual page

3.10. Preventing kernel modules from being automatically loaded at system boot time

You can prevent the system from loading a kernel module automatically during the boot process by listing the module in modprobe configuration file with a corresponding command.

Prerequisites

  • The commands in this procedure require root privileges. Either use su - to switch to the root user or preface the commands with sudo.
  • The kmod package is installed.
  • Ensure that your current system configuration does not require a kernel module you plan to deny.

Procedure

  1. List modules loaded to the currently running kernel by using the lsmod command:

    $ lsmod
    Module                  Size  Used by
    tls                   131072  0
    uinput                 20480  1
    snd_seq_dummy          16384  0
    snd_hrtimer            16384  1
    …

    In the output, identify the module you want to prevent from being loaded.

    • Alternatively, identify an unloaded kernel module you want to prevent from potentially loading in the /lib/modules/<KERNEL-VERSION>/kernel/<SUBSYSTEM>/ directory, for example:

      $ ls /lib/modules/4.18.0-477.20.1.el8_8.x86_64/kernel/crypto/
      ansi_cprng.ko.xz        chacha20poly1305.ko.xz  md4.ko.xz               serpent_generic.ko.xz
      anubis.ko.xz            cmac.ko.xz…
  2. Create a configuration file serving as a denylist:

    # touch /etc/modprobe.d/denylist.conf
  3. In a text editor of your choice, combine the names of modules you want to exclude from automatic loading to the kernel with the blacklist configuration command, for example:

    # Prevents <KERNEL-MODULE-1> from being loaded
    blacklist <MODULE-NAME-1>
    install <MODULE-NAME-1> /bin/false
    
    # Prevents <KERNEL-MODULE-2> from being loaded
    blacklist <MODULE-NAME-2>
    install <MODULE-NAME-2> /bin/false
    …

    Because the blacklist command does not prevent the module from being loaded as a dependency for another kernel module that is not in a denylist, you must also define the install line. In this case, the system runs /bin/false instead of installing the module. The lines starting with a hash sign are comments you can use to make the file more readable.

    Note

    When entering the name of a kernel module, do not append the .ko.xz extension to the end of the name. Kernel module names do not have extensions; their corresponding files do.

  4. Create a backup copy of the current initial RAM disk image before rebuilding:

    # cp /boot/initramfs-$(uname -r).img /boot/initramfs-$(uname -r).bak.$(date +%m-%d-%H%M%S).img
    • Alternatively, create a backup copy of an initial RAM disk image which corresponds to the kernel version for which you want to prevent kernel modules from automatic loading:

      # cp /boot/initramfs-<VERSION>.img /boot/initramfs-<VERSION>.img.bak.$(date +%m-%d-%H%M%S)
  5. Generate a new initial RAM disk image to apply the changes:

    # dracut -f -v
    • If you build an initial RAM disk image for a different kernel version than your system currently uses, specify both target initramfs and kernel version:

      # dracut -f -v /boot/initramfs-<TARGET-VERSION>.img <CORRESPONDING-TARGET-KERNEL-VERSION>
  6. Restart the system:

    $ reboot
Important

The changes described in this procedure will take effect and persist after rebooting the system. If you incorrectly list a key kernel module in the denylist, you can switch the system to an unstable or non-operational state.

Additional resources

3.11. Compiling custom kernel modules

You can build a sampling kernel module as requested by various configurations at hardware and software level.

Prerequisites

  • You installed the kernel-devel, gcc, and elfutils-libelf-devel packages.

    # dnf install kernel-devel-$(uname -r) gcc elfutils-libelf-devel
  • You have root permissions.
  • You created the /root/testmodule/ directory where you compile the custom kernel module.

Procedure

  1. Create the /root/testmodule/test.c file with the following content.

    #include <linux/module.h>
    #include <linux/kernel.h>
    
    int init_module(void)
        { printk("Hello World\n This is a test\n"); return 0; }
    
    void cleanup_module(void)
        { printk("Good Bye World"); }
    
    MODULE_LICENSE("GPL");

    The test.c file is a source file that provides the main functionality to the kernel module. The file has been created in a dedicated /root/testmodule/ directory for organizational purposes. After the module compilation, the /root/testmodule/ directory will contain multiple files.

    The test.c file includes from the system libraries:

    • The linux/kernel.h header file is necessary for the printk() function in the example code.
    • The linux/module.h file contains function declarations and macro definitions to be shared between several source files written in C programming language.
  2. Follow the init_module() and cleanup_module() functions to start and end the kernel logging function printk(), which prints text.
  3. Create the /root/testmodule/Makefile file with the following content.

    obj-m := test.o

    The Makefile contains instructions that the compiler has to produce an object file specifically named test.o. The obj-m directive specifies that the resulting test.ko file is going to be compiled as a loadable kernel module. Alternatively, the obj-y directive would instruct to build test.ko as a built-in kernel module.

  4. Compile the kernel module.

    # make -C /lib/modules/$(uname -r)/build M=/root/testmodule modules
    make: Entering directory '/usr/src/kernels/5.14.0-70.17.1.el9_0.x86_64'
      CC [M]  /root/testmodule/test.o
      MODPOST /root/testmodule/Module.symvers
      CC [M]  /root/testmodule/test.mod.o
      LD [M]  /root/testmodule/test.ko
      BTF [M] /root/testmodule/test.ko
    Skipping BTF generation for /root/testmodule/test.ko due to unavailability of vmlinux
    make: Leaving directory '/usr/src/kernels/5.14.0-70.17.1.el9_0.x86_64'

    The compiler creates an object file (test.o) for each source file (test.c) as an intermediate step before linking them together into the final kernel module (test.ko).

    After a successful compilation, /root/testmodule/ contains additional files that relate to the compiled custom kernel module. The compiled module itself is represented by the test.ko file.

Verification

  1. Optional: check the contents of the /root/testmodule/ directory:

    # ls -l /root/testmodule/
    total 152
    -rw-r—​r--. 1 root root    16 Jul 26 08:19 Makefile
    -rw-r—​r--. 1 root root    25 Jul 26 08:20 modules.order
    -rw-r—​r--. 1 root root     0 Jul 26 08:20 Module.symvers
    -rw-r—​r--. 1 root root   224 Jul 26 08:18 test.c
    -rw-r—​r--. 1 root root 62176 Jul 26 08:20 test.ko
    -rw-r—​r--. 1 root root    25 Jul 26 08:20 test.mod
    -rw-r—​r--. 1 root root   849 Jul 26 08:20 test.mod.c
    -rw-r—​r--. 1 root root 50936 Jul 26 08:20 test.mod.o
    -rw-r—​r--. 1 root root 12912 Jul 26 08:20 test.o
  2. Copy the kernel module to the /lib/modules/$(uname -r)/ directory:

    # cp /root/testmodule/test.ko /lib/modules/$(uname -r)/
  3. Update the modular dependency list:

    # depmod -a
  4. Load the kernel module:

    # modprobe -v test
    insmod /lib/modules/5.14.0-1.el9.x86_64/test.ko
  5. Verify that the kernel module was successfully loaded:

    # lsmod | grep test
    test                   16384  0
  6. Read the latest messages from the kernel ring buffer:

    # dmesg
    [74422.545004] Hello World
                    This is a test

Additional resources

Chapter 4. Configuring kernel command-line parameters

With kernel command-line parameters, you can change the behavior of certain aspects of the Red Hat Enterprise Linux kernel at boot time. As a system administrator, you have full control over what options get set at boot. Certain kernel behaviors can only be set at boot time, so understanding how to make these changes is a key administration skill.

Important

Changing the behavior of the system by modifying kernel command-line parameters may have negative effects on your system. Always test changes prior to deploying them in production. For further guidance, contact Red Hat Support.

4.1. What are kernel command-line parameters

With kernel command-line parameters, you can overwrite default values and set specific hardware settings. At boot time, you can configure the following features:

  • The Red Hat Enterprise Linux kernel
  • The initial RAM disk
  • The user space features

By default, the kernel command-line parameters for systems using the GRUB boot loader are defined in the boot entry configuration file for each kernel boot entry.

You can manipulate boot loader configuration files by using the grubby utility. With grubby, you can perform these actions:

  • Change the default boot entry.
  • Add or remove arguments from a GRUB menu entry.

Additional resources

4.2. Understanding boot entries

A boot entry is a collection of options which are stored in a configuration file and tied to a particular kernel version. In practice, you have at least as many boot entries as your system has installed kernels. The boot entry configuration file is located in the /boot/loader/entries/ directory and can look like this:

d8712ab6d4f14683c5625e87b52b6b6e-5.14.0-1.el9.x86_64.conf

The file name above consists of a machine ID stored in the /etc/machine-id file, and a kernel version.

The boot entry configuration file contains information about the kernel version, the initial ramdisk image, and the kernel command-line parameters. The example contents of a boot entry config can be seen below:

title Red Hat Enterprise Linux (5.14.0-1.el9.x86_64) 9.0 (Plow)
version 5.14.0-1.el9.x86_64
linux /vmlinuz-5.14.0-1.el9.x86_64
initrd /initramfs-5.14.0-1.el9.x86_64.img
options root=/dev/mapper/rhel_kvm--02--guest08-root ro crashkernel=1G-4G:192M,4G-64G:256M,64G-:512M resume=/dev/mapper/rhel_kvm--02--guest08-swap rd.lvm.lv=rhel_kvm-02-guest08/root rd.lvm.lv=rhel_kvm-02-guest08/swap console=ttyS0,115200
grub_users $grub_users
grub_arg --unrestricted
grub_class kernel

4.3. Changing kernel command-line parameters for all boot entries

Change kernel command-line parameters for all boot entries on your system.

Important

When installing a newer version of the kernel in RHEL 9 systems, the grubby tool passes the kernel command-line arguments from the previous kernel version.

However, this does not apply to RHEL version 9.0 in which newly installed kernels lose previous command-line options. You must run the grub2-mkconfig command on the newly installed kernel to pass the parameters to your new kernel. For more information about this known issue, see Boot loader.

Prerequisites

  • Verify that the grubby utility is installed on your system.
  • Verify that the zipl utility is installed on your IBM Z system.

Procedure

  • To add a parameter:

    # grubby --update-kernel=ALL --args="<NEW_PARAMETER>"

    For systems that use the GRUB boot loader and, on IBM Z that use the zIPL boot loader, the command adds a new kernel parameter to each /boot/loader/entries/<ENTRY>.conf file.

    • On IBM Z, update the boot menu:

      # zipl
  • To remove a parameter:

    # grubby --update-kernel=ALL --remove-args="<PARAMETER_TO_REMOVE>"
    • On IBM Z, update the boot menu:

      # zipl

Additional resources

4.4. Changing kernel command-line parameters for a single boot entry

Make changes in kernel command-line parameters for a single boot entry on your system.

Prerequisites

  • Verify that the grubby and zipl utilities are installed on your system.

Procedure

  • To add a parameter:

    # grubby --update-kernel=/boot/vmlinuz-$(uname -r) --args="<NEW_PARAMETER>"
    • On IBM Z, update the boot menu:

      # zipl
  • To remove a parameter:

    # grubby --update-kernel=/boot/vmlinuz-$(uname -r) --remove-args="<PARAMETER_TO_REMOVE>"
    • On IBM Z, update the boot menu:

      # zipl
Important
  • grubby modifies and stores the kernel command-line parameters of an individual kernel boot entry in the /boot/loader/entries/<ENTRY>.conf file.

Additional resources

4.5. Changing kernel command-line parameters temporarily at boot time

Make temporary changes to a Kernel Menu Entry by changing the kernel parameters only during a single boot process.

Note

This procedure applies only for a single boot and does not persistently make the changes.

Procedure

  1. Boot into the GRUB 2 boot menu.
  2. Select the kernel you want to start.
  3. Press the e key to edit the kernel parameters.
  4. Find the kernel command line by moving the cursor down. The kernel command line starts with linux on 64-Bit IBM Power Series and x86-64 BIOS-based systems, or linuxefi on UEFI systems.
  5. Move the cursor to the end of the line.

    Note

    Press Ctrl+a to jump to the start of the line and Ctrl+e to jump to the end of the line. On some systems, Home and End keys might also work.

  6. Edit the kernel parameters as required. For example, to run the system in emergency mode, add the emergency parameter at the end of the linux line:

    linux   ($root)/vmlinuz-5.14.0-63.el9.x86_64 root=/dev/mapper/rhel-root ro crashkernel=1G-4G:192M,4G-64G:256M,64G-:512M resume=/dev/mapper/rhel-swap rd.lvm.lv=rhel/root rd.lvm.lv=rhel/swap rhgb quiet emergency

    To enable the system messages, remove the rhgb and quiet parameters.

  7. Press Ctrl+x to boot with the selected kernel and the modified command line parameters.
Important

If you press the Esc key to leave command line editing, it will drop all the user made changes.

4.6. Configuring GRUB settings to enable serial console connection

The serial console is beneficial when you need to connect to a headless server or an embedded system and the network is down. Or when you need to avoid security rules and obtain login access on a different system.

You need to configure some default GRUB settings to use the serial console connection.

Prerequisites

  • You have root permissions.

Procedure

  1. Add the following two lines to the /etc/default/grub file:

    GRUB_TERMINAL="serial"
    GRUB_SERIAL_COMMAND="serial --speed=9600 --unit=0 --word=8 --parity=no --stop=1"

    The first line disables the graphical terminal. The GRUB_TERMINAL key overrides values of GRUB_TERMINAL_INPUT and GRUB_TERMINAL_OUTPUT keys.

    The second line adjusts the baud rate (--speed), parity and other values to fit your environment and hardware. Note that a much higher baud rate, for example 115200, is preferable for tasks such as following log files.

  2. Update the GRUB configuration file.

    • On BIOS-based machines:

      # grub2-mkconfig -o /boot/grub2/grub.cfg
    • On UEFI-based machines:

      # grub2-mkconfig -o /boot/grub2/grub.cfg
  3. Reboot the system for the changes to take effect.

4.7. Changing boot entries with the GRUB configuration file

The /etc/default/grub GRUB configuration file contains the GRUB_CMDLINE_LINUX key, which lists kernel command-line arguments to add to boot entries for the Linux kernel. For example:

GRUB_CMDLINE_LINUX="crashkernel=1G-4G:192M,4G-64G:256M,64G-:512M resume=/dev/mapper/rhel-swap rd.lvm.lv=rhel/root rd.lvm.lv=rhel/swap"

To change the boot entries, overwrite Boot Loader Specification (BLS) snippets with the contents of the GRUB_CMDLINE_LINUX values.

Prerequisites

  • A fresh RHEL 9 installation.

Procedure

  1. Add or remove a kernel parameter for individual kernels in a post installation script with grubby:

    # grubby --update-kernel <PATH_TO_KERNEL> --args "<NEW_ARGUMENTS>"

    For example, add the noapic parameter to the chosen kernel:

    # grubby --update-kernel /boot/vmlinuz-5.14.0-362.8.1.el9_3.x86_64 --args "noapic"

    The parameter is propagated into the BLS snippets, but not into the /etc/default/grub file.

  2. Overwrite BLS snippets with the contents of the GRUB_CMDLINE_LINUX values present in the /etc/default/grub file:

    # grub2-mkconfig -o /boot/grub2/grub.cfg --update-bls-cmdline
    Generating grub configuration file …​
    Adding boot menu entry for UEFI Firmware Settings …​
    done
    Note

    Other changes, such as changes made to GRUB_TIMEOUT key (also included in the /etc/default/grub GRUB configuration file), do get propagated to the new grub.cfg by default.

Verification

  1. Reboot your operating system.
  2. Verify that the parameters are included in the /proc/cmdline file.

    For example, /proc/cmdline contains the noapic kernel parameter:

    BOOT_IMAGE=(hd0,gpt2)/vmlinuz-4.18.0-425.3.1.el8.x86_64 root=/dev/mapper/RHELCSB-Root ro vconsole.keymap=us crashkernel=auto rd.lvm.lv=RHELCSB/Root rd.luks.uuid=luks-d8a28c4c-96aa-4319-be26-96896272151d rhgb quiet noapic rd.luks.key=d8a28c4c-96aa-4319-be26-96896272151d=/keyfile:UUID=c47d962e-4be8-41d6-8216-8cf7a0d3b911 ipv6.disable=1

Chapter 5. Configuring kernel parameters at runtime

As a system administrator, you can modify many facets of the Red Hat Enterprise Linux kernel’s behavior at runtime. Configure kernel parameters at runtime by using the sysctl command and by modifying the configuration files in the /etc/sysctl.d/ and /proc/sys/ directories.

Important

Configuring kernel parameters on a production system requires careful planning. Unplanned changes may render the kernel unstable, requiring a system reboot. Verify that you are using valid options before changing any kernel values.

5.1. What are kernel parameters

Kernel parameters are tunable values which you can adjust while the system is running. There is no requirement to reboot or recompile the kernel for changes to take effect.

It is possible to address the kernel parameters through:

  • The sysctl command
  • The virtual file system mounted at the /proc/sys/ directory
  • The configuration files in the /etc/sysctl.d/ directory

Tunables are divided into classes by the kernel subsystem. Red Hat Enterprise Linux has the following tunable classes:

Table 5.1. Table of sysctl classes

Tunable classSubsystem

abi

Execution domains and personalities

crypto

Cryptographic interfaces

debug

Kernel debugging interfaces

dev

Device-specific information

fs

Global and specific file system tunables

kernel

Global kernel tunables

net

Network tunables

sunrpc

Sun Remote Procedure Call (NFS)

user

User Namespace limits

vm

Tuning and management of memory, buffers, and cache

Additional resources

  • sysctl(8), and sysctl.d(5) manual pages

5.2. Configuring kernel parameters temporarily with sysctl

Use the sysctl command to temporarily set kernel parameters at runtime. The command is also useful for listing and filtering tunables.

Prerequisites

  • Root permissions

Procedure

  1. List all parameters and their values.

    # sysctl -a
    Note

    The # sysctl -a command displays kernel parameters, which can be adjusted at runtime and at boot time.

  2. To configure a parameter temporarily, enter:

    # sysctl <TUNABLE_CLASS>.<PARAMETER>=<TARGET_VALUE>

    The sample command above changes the parameter value while the system is running. The changes take effect immediately, without a need for restart.

    Note

    The changes return back to default after your system reboots.

5.3. Configuring kernel parameters permanently with sysctl

Use the sysctl command to permanently set kernel parameters.

Prerequisites

  • Root permissions

Procedure

  1. List all parameters.

    # sysctl -a

    The command displays all kernel parameters that can be configured at runtime.

  2. Configure a parameter permanently:

    # sysctl -w <TUNABLE_CLASS>.<PARAMETER>=<TARGET_VALUE> >> /etc/sysctl.conf

    The sample command changes the tunable value and writes it to the /etc/sysctl.conf file, which overrides the default values of kernel parameters. The changes take effect immediately and persistently, without a need for restart.

Note

To permanently modify kernel parameters you can also make manual changes to the configuration files in the /etc/sysctl.d/ directory.

Additional resources

5.4. Using configuration files in /etc/sysctl.d/ to adjust kernel parameters

Modify configuration files in the /etc/sysctl.d/ directory manually to permanently set kernel parameters.

Prerequisites

  • Root permissions

Procedure

  1. Create a new configuration file in /etc/sysctl.d/.

    # vim /etc/sysctl.d/<some_file.conf>
  2. Include kernel parameters, one per line.

    <TUNABLE_CLASS>.<PARAMETER>=<TARGET_VALUE>
    <TUNABLE_CLASS>.<PARAMETER>=<TARGET_VALUE>
  3. Save the configuration file.
  4. Reboot the machine for the changes to take effect.

    • Alternatively, to apply changes without rebooting, enter:

      # sysctl -p /etc/sysctl.d/<some_file.conf>

      The command enables you to read values from the configuration file, which you created earlier.

Additional resources

  • sysctl(8), sysctl.d(5) manual pages

5.5. Configuring kernel parameters temporarily through /proc/sys/

Set kernel parameters temporarily through the files in the /proc/sys/ virtual file system directory.

Prerequisites

  • Root permissions

Procedure

  1. Identify a kernel parameter you want to configure.

    # ls -l /proc/sys/<TUNABLE_CLASS>/

    The writable files returned by the command can be used to configure the kernel. The files with read-only permissions provide feedback on the current settings.

  2. Assign a target value to the kernel parameter.

    # echo <TARGET_VALUE> > /proc/sys/<TUNABLE_CLASS>/<PARAMETER>

    The command makes configuration changes that will disappear once the system is restarted.

  3. Optionally, verify the value of the newly set kernel parameter.

    # cat /proc/sys/<TUNABLE_CLASS>/<PARAMETER>

5.6. Additional resources

Chapter 6. Configuring kernel parameters permanently by using the RHEL system role

You can use the kernel_settings RHEL system role to configure kernel parameters on multiple clients at once. This solution:

  • Provides a friendly interface with efficient input setting.
  • Keeps all intended kernel parameters in one place.

After you run the kernel_settings role from the control machine, the kernel parameters are applied to the managed systems immediately and persist across reboots.

Important

Note that RHEL system role delivered over RHEL channels are available to RHEL customers as an RPM package in the default AppStream repository. RHEL system role are also available as a collection to customers with Ansible subscriptions over Ansible Automation Hub.

6.1. Introduction to the kernel_settings RHEL system role

RHEL system roles is a set of roles that provide a consistent configuration interface to remotely manage multiple systems.

RHEL system roles were introduced for automated configurations of the kernel using the kernel_settings RHEL system role. The rhel-system-roles package contains this system role, and also the reference documentation.

To apply the kernel parameters on one or more systems in an automated fashion, use the kernel_settings role with one or more of its role variables of your choice in a playbook. A playbook is a list of one or more plays that are human-readable, and are written in the YAML format.

You can use an inventory file to define a set of systems that you want Ansible to configure according to the playbook.

With the kernel_settings role you can configure:

  • The kernel parameters using the kernel_settings_sysctl role variable
  • Various kernel subsystems, hardware devices, and device drivers using the kernel_settings_sysfs role variable
  • The CPU affinity for the systemd service manager and processes it forks using the kernel_settings_systemd_cpu_affinity role variable
  • The kernel memory subsystem transparent hugepages using the kernel_settings_transparent_hugepages and kernel_settings_transparent_hugepages_defrag role variables

Additional resources

6.2. Applying selected kernel parameters by using the kernel_settings RHEL system role

Follow these steps to prepare and apply an Ansible playbook to remotely configure kernel parameters with persisting effect on multiple managed operating systems.

Prerequisites

Procedure

  1. Create a playbook file, for example ~/playbook.yml, with the following content:

    ---
    - name: Configure kernel settings
      hosts: managed-node-01.example.com
      roles:
        - rhel-system-roles.kernel_settings
      vars:
        kernel_settings_sysctl:
          - name: fs.file-max
            value: 400000
          - name: kernel.threads-max
            value: 65536
        kernel_settings_sysfs:
          - name: /sys/class/net/lo/mtu
            value: 65000
        kernel_settings_transparent_hugepages: madvise
    • name: optional key which associates an arbitrary string with the play as a label and identifies what the play is for.
    • hosts: key in the play which specifies the hosts against which the play is run. The value or values for this key can be provided as individual names of managed hosts or as groups of hosts as defined in the inventory file.
    • vars: section of the playbook which represents a list of variables containing selected kernel parameter names and values to which they have to be set.
    • role: key which specifies what RHEL system role is going to configure the parameters and values mentioned in the vars section.

      Note

      You can modify the kernel parameters and their values in the playbook to fit your needs.

  2. Validate the playbook syntax:

    $ ansible-playbook --syntax-check ~/playbook.yml

    Note that this command only validates the syntax and does not protect against a wrong but valid configuration.

  3. Run the playbook:

    $ ansible-playbook ~/playbook.yml
  4. Restart your managed hosts and check the affected kernel parameters to verify that the changes have been applied and persist across reboots.

Additional resources

Chapter 7. Applying patches with kernel live patching

You can use the Red Hat Enterprise Linux kernel live patching solution to patch a running kernel without rebooting or restarting any processes.

With this solution, system administrators:

  • Can immediately apply critical security patches to the kernel.
  • Do not have to wait for long-running tasks to complete, for users to log off, or for scheduled downtime.
  • Control the system’s uptime more and do not sacrifice security or stability.

Note that not every critical or important CVE will be resolved using the kernel live patching solution. Our goal is to reduce the required reboots for security-related patches, not to eliminate them entirely. For more details about the scope of live patching, see the Customer Portal Solutions article.

Warning

Some incompatibilities exist between kernel live patching and other kernel subcomponents. Read the

Limitations of kpatch carefully before using kernel live patching.

Note

For details about the support cadence of kernel live patching updates, see:

7.1. Limitations of kpatch

  • The kpatch feature is not a general-purpose kernel upgrade mechanism. It is used for applying simple security and bug fix updates when rebooting the system is not immediately possible.
  • Do not use the SystemTap or kprobe tools during or after loading a patch. The patch could fail to take effect until after such probes have been removed.

7.2. Support for third-party live patching

The kpatch utility is the only kernel live patching utility supported by Red Hat with the RPM modules provided by Red Hat repositories. Red Hat will not support any live patches which were not provided by Red Hat itself.

If you require support for an issue that arises with a third-party live patch, Red Hat recommends that you open a case with the live patching vendor at the outset of any investigation in which a root cause determination is necessary. This allows the source code to be supplied if the vendor allows, and for their support organization to provide assistance in root cause determination prior to escalating the investigation to Red Hat Support.

For any system running with third-party live patches, Red Hat reserves the right to ask for reproduction with Red Hat shipped and supported software. In the event that this is not possible, we require a similar system and workload be deployed on your test environment without live patches applied, to confirm if the same behavior is observed.

For more information about third-party software support policies, see How does Red Hat Global Support Services handle third-party software, drivers, and/or uncertified hardware/hypervisors or guest operating systems?

7.3. Access to kernel live patches

Kernel live patching capability is implemented as a kernel module (kmod) that is delivered as an RPM package.

All customers have access to kernel live patches, which are delivered through the usual channels. However, customers who do not subscribe to an extended support offering will lose access to new patches for the current minor release once the next minor release becomes available. For example, customers with standard subscriptions will only be able to live patch RHEL 9.1 kernel until the RHEL 9.2 kernel is released.

7.4. Components of kernel live patching

The components of kernel live patching are as follows:

Kernel patch module

  • The delivery mechanism for kernel live patches.
  • A kernel module which is built specifically for the kernel being patched.
  • The patch module contains the code of the desired fixes for the kernel.
  • The patch modules register with the livepatch kernel subsystem and provide information about original functions to be replaced, with corresponding pointers to the replacement functions. Kernel patch modules are delivered as RPMs.
  • The naming convention is kpatch_<kernel version>_<kpatch version>_<kpatch release>. The "kernel version" part of the name has dots replaced with underscores.
The kpatch utility
A command-line utility for managing patch modules.
The kpatch service
A systemd service required by multiuser.target. This target loads the kernel patch module at boot time.
The kpatch-dnf package
A DNF plugin delivered in the form of an RPM package. This plugin manages automatic subscription to kernel live patches.

7.5. How kernel live patching works

The kpatch kernel patching solution uses the livepatch kernel subsystem to redirect old functions to new ones. When a live kernel patch is applied to a system, the following things happen:

  1. The kernel patch module is copied to the /var/lib/kpatch/ directory and registered for re-application to the kernel by systemd on next boot.
  2. The kpatch module is loaded into the running kernel and the new functions are registered to the ftrace mechanism with a pointer to the location in memory of the new code.
  3. When the kernel accesses the patched function, it is redirected by the ftrace mechanism which bypasses the original functions and redirects the kernel to patched version of the function.

Figure 7.1. How kernel live patching works

rhel kpatch overview

7.6. Subscribing the currently installed kernels to the live patching stream

A kernel patch module is delivered in an RPM package, specific to the version of the kernel being patched. Each RPM package will be cumulatively updated over time.

The following procedure explains how to subscribe to all future cumulative live patching updates for a given kernel. Because live patches are cumulative, you cannot select which individual patches are deployed for a given kernel.

Warning

Red Hat does not support any third party live patches applied to a Red Hat supported system.

Prerequisites

  • Root permissions

Procedure

  1. Optionally, check your kernel version:

    # uname -r
    5.14.0-1.el9.x86_64
  2. Search for a live patching package that corresponds to the version of your kernel:

    # dnf search $(uname -r)
  3. Install the live patching package:

    # dnf install "kpatch-patch = $(uname -r)"

    The command above installs and applies the latest cumulative live patches for that specific kernel only.

    If the version of a live patching package is 1-1 or higher, the package will contain a patch module. In that case the kernel will be automatically patched during the installation of the live patching package.

    The kernel patch module is also installed into the /var/lib/kpatch/ directory to be loaded by the systemd system and service manager during the future reboots.

    Note

    An empty live patching package will be installed when there are no live patches available for a given kernel. An empty live patching package will have a kpatch_version-kpatch_release of 0-0, for example kpatch-patch-5_14_0-1-0-0.x86_64.rpm. The installation of the empty RPM subscribes the system to all future live patches for the given kernel.

Verification

  • Verify that all installed kernels have been patched:

    # kpatch list
    Loaded patch modules:
    kpatch_5_14_0_1_0_1 [enabled]
    
    Installed patch modules:
    kpatch_5_14_0_1_0_1 (5.14.0-1.el9.x86_64)
    …​

    The output shows that the kernel patch module has been loaded into the kernel that is now patched with the latest fixes from the kpatch-patch-5_14_0-1-0-1.el9.x86_64.rpm package.

    Note

    Entering the kpatch list command does not return an empty live patching package. Use the rpm -qa | grep kpatch command instead.

    # rpm -qa | grep kpatch
    kpatch-dnf-0.4-3.el9.noarch
    kpatch-0.9.7-2.el9.noarch
    kpatch-patch-5_14_0-284_25_1-0-0.el9_2.x86_64

Additional resources

7.7. Automatically subscribing any future kernel to the live patching stream

You can use the kpatch-dnf DNF plugin to subscribe your system to fixes delivered by the kernel patch module, also known as kernel live patches. The plugin enables automatic subscription for any kernel the system currently uses, and also for kernels to-be-installed in the future.

Prerequisites

  • You have root permissions.

Procedure

  1. Optionally, check all installed kernels and the kernel you are currently running:

    # dnf list installed | grep kernel
    Updating Subscription Management repositories.
    Installed Packages
    ...
    kernel-core.x86_64            5.14.0-1.el9              @beaker-BaseOS
    kernel-core.x86_64            5.14.0-2.el9              @@commandline
    ...
    
    # uname -r
    5.14.0-2.el9.x86_64
  2. Install the kpatch-dnf plugin:

    # dnf install kpatch-dnf
  3. Enable automatic subscription to kernel live patches:

    # dnf kpatch auto
    Updating Subscription Management repositories.
    Last metadata expiration check: 1:38:21 ago on Fri 17 Sep 2021 07:29:53 AM EDT.
    Dependencies resolved.
    ==================================================
     Package                             Architecture
    ==================================================
    Installing:
     kpatch-patch-5_14_0-1               x86_64
     kpatch-patch-5_14_0-2               x86_64
    
    Transaction Summary
    ===================================================
    Install  2 Packages
    …​

    This command subscribes all currently installed kernels to receiving kernel live patches. The command also installs and applies the latest cumulative live patches, if any, for all installed kernels.

    In the future, when you update the kernel, live patches will automatically be installed during the new kernel installation process.

    The kernel patch module is also installed into the /var/lib/kpatch/ directory to be loaded by the systemd system and service manager during future reboots.

    Note

    An empty live patching package will be installed when there are no live patches available for a given kernel. An empty live patching package will have a kpatch_version-kpatch_release of 0-0, for example kpatch-patch-5_14_0-1-0-0.el9.x86_64.rpm. The installation of the empty RPM subscribes the system to all future live patches for the given kernel.

Verification

  • Verify that all installed kernels have been patched:

    # kpatch list
    Loaded patch modules:
    kpatch_5_14_0_2_0_1 [enabled]
    
    Installed patch modules:
    kpatch_5_14_0_1_0_1 (5.14.0-1.el9.x86_64)
    kpatch_5_14_0_2_0_1 (5.14.0-2.el9.x86_64)

    The output shows that both the kernel you are running, and the other installed kernel have been patched with fixes from kpatch-patch-5_14_0-1-0-1.el9.x86_64.rpm and kpatch-patch-5_14_0-2-0-1.el9.x86_64.rpm packages respectively.

    Note

    Entering the kpatch list command does not return an empty live patching package. Use the rpm -qa | grep kpatch command instead.

    # rpm -qa | grep kpatch
    kpatch-dnf-0.4-3.el9.noarch
    kpatch-0.9.7-2.el9.noarch
    kpatch-patch-5_14_0-284_25_1-0-0.el9_2.x86_64

Additional resources

  • kpatch(1) and dnf-kpatch(8) manual pages

7.8. Disabling automatic subscription to the live patching stream

When you subscribe your system to fixes delivered by the kernel patch module, your subscription is automatic. You can disable this feature, and thus disable automatic installation of kpatch-patch packages.

Prerequisites

  • You have root permissions.

Procedure

  1. Optionally, check all installed kernels and the kernel you are currently running:

    # dnf list installed | grep kernel
    Updating Subscription Management repositories.
    Installed Packages
    ...
    kernel-core.x86_64            5.14.0-1.el9              @beaker-BaseOS
    kernel-core.x86_64            5.14.0-2.el9              @@commandline
    ...
    
    # uname -r
    5.14.0-2.el9.x86_64
  2. Disable automatic subscription to kernel live patches:

    # dnf kpatch manual
    Updating Subscription Management repositories.

Verification step

  • You can check for the successful outcome:

    # yum kpatch status
    ...
    Updating Subscription Management repositories.
    Last metadata expiration check: 0:30:41 ago on Tue Jun 14 15:59:26 2022.
    Kpatch update setting: manual

Additional resources

  • kpatch(1) and dnf-kpatch(8) manual pages

7.9. Updating kernel patch modules

Since kernel patch modules are delivered and applied through RPM packages, updating a cumulative kernel patch module is like updating any other RPM package.

Prerequisites

Procedure

  • Update to a new cumulative version for the current kernel:

    # dnf update "kpatch-patch = $(uname -r)"

    The command above automatically installs and applies any updates that are available for the currently running kernel. Including any future released cumulative live patches.

  • Alternatively, update all installed kernel patch modules:

    # dnf update "kpatch-patch"
Note

When the system reboots into the same kernel, the kernel is automatically live patched again by the kpatch.service systemd service.

Additional resources

7.10. Removing the live patching package

Disable the Red Hat Enterprise Linux kernel live patching solution by removing the live patching package.

Prerequisites

  • Root permissions
  • The live patching package is installed.

Procedure

  1. Select the live patching package.

    # dnf list installed | grep kpatch-patch
    kpatch-patch-5_14_0-1.x86_64        0-1.el9        @@commandline
    …​

    The example output above lists live patching packages that you installed.

  2. Remove the live patching package.

    # dnf remove kpatch-patch-5_14_0-1.x86_64

    When a live patching package is removed, the kernel remains patched until the next reboot, but the kernel patch module is removed from disk. On future reboot, the corresponding kernel will no longer be patched.

  3. Reboot your system.
  4. Verify that the live patching package has been removed.

    # dnf list installed | grep kpatch-patch

    The command displays no output if the package has been successfully removed.

  5. Optionally, verify that the kernel live patching solution is disabled.

    # kpatch list
    Loaded patch modules:

    The example output shows that the kernel is not patched and the live patching solution is not active because there are no patch modules that are currently loaded.

Important

Currently, Red Hat does not support reverting live patches without rebooting your system. In case of any issues, contact our support team.

Additional resources

7.11. Uninstalling the kernel patch module

Prevent the Red Hat Enterprise Linux kernel live patching solution from applying a kernel patch module on subsequent boots.

Prerequisites

  • Root permissions
  • A live patching package is installed.
  • A kernel patch module is installed and loaded.

Procedure

  1. Select a kernel patch module:

    # kpatch list
    Loaded patch modules:
    kpatch_5_14_0_1_0_1 [enabled]
    
    Installed patch modules:
    kpatch_5_14_0_1_0_1 (5.14.0-1.el9.x86_64)
    …​
  2. Uninstall the selected kernel patch module.

    # kpatch uninstall kpatch_5_14_0_1_0_1
    uninstalling kpatch_5_14_0_1_0_1 (5.14.0-1.el9.x86_64)
    • Note that the uninstalled kernel patch module is still loaded:

      # kpatch list
      Loaded patch modules:
      kpatch_5_14_0_1_0_1 [enabled]
      
      Installed patch modules:
      <NO_RESULT>

      When the selected module is uninstalled, the kernel remains patched until the next reboot, but the kernel patch module is removed from disk.

  3. Reboot your system.
  4. Optionally, verify that the kernel patch module has been uninstalled.

    # kpatch list
    Loaded patch modules:
    …​

    The example output above shows no loaded or installed kernel patch modules, therefore the kernel is not patched and the kernel live patching solution is not active.

Important

Currently, Red Hat does not support reverting live patches without rebooting your system. In case of any issues, contact our support team.

Additional resources

  • The kpatch(1) manual page

7.12. Disabling kpatch.service

Prevent the Red Hat Enterprise Linux kernel live patching solution from applying all kernel patch modules globally on subsequent boots.

Prerequisites

  • Root permissions
  • A live patching package is installed.
  • A kernel patch module is installed and loaded.

Procedure

  1. Verify kpatch.service is enabled.

    # systemctl is-enabled kpatch.service
    enabled
  2. Disable kpatch.service:

    # systemctl disable kpatch.service
    Removed /etc/systemd/system/multi-user.target.wants/kpatch.service.
    • Note that the applied kernel patch module is still loaded:

      # kpatch list
      Loaded patch modules:
      kpatch_5_14_0_1_0_1 [enabled]
      
      Installed patch modules:
      kpatch_5_14_0_1_0_1 (5.14.0-1.el9.x86_64)
  3. Reboot your system.
  4. Optionally, verify the status of kpatch.service.

    # systemctl status kpatch.service
    ● kpatch.service - "Apply kpatch kernel patches"
       Loaded: loaded (/usr/lib/systemd/system/kpatch.service; disabled; vendor preset: disabled)
       Active: inactive (dead)

    The example output testifies that kpatch.service has been disabled and is not running. Thereby, the kernel live patching solution is not active.

  5. Verify that the kernel patch module has been unloaded.

    # kpatch list
    Loaded patch modules:
    
    Installed patch modules:
    kpatch_5_14_0_1_0_1 (5.14.0-1.el9.x86_64)

    The example output above shows that a kernel patch module is still installed but the kernel is not patched.

Important

Currently, Red Hat does not support reverting live patches without rebooting your system. In case of any issues, contact our support team.

Additional resources

Chapter 8. Keeping kernel panic parameters disabled in virtualized environments

When configuring a Virtual Machine in RHEL 9, you should not enable the softlockup_panic and nmi_watchdog kernel parameters, because the Virtual Machine might suffer from a spurious soft lockup. And that should not require a kernel panic.

Find the reasons behind this advice in the following sections.

8.1. What is a soft lockup

A soft lockup is a situation usually caused by a bug, when a task is executing in kernel space on a CPU without rescheduling. The task also does not allow any other task to execute on that particular CPU. As a result, a warning is displayed to a user through the system console. This problem is also referred to as the soft lockup firing.

Additional resources

8.2. Parameters controlling kernel panic

The following kernel parameters can be set to control a system’s behavior when a soft lockup is detected.

softlockup_panic

Controls whether or not the kernel will panic when a soft lockup is detected.

TypeValueEffect

Integer

0

kernel does not panic on soft lockup

Integer

1

kernel panics on soft lockup

By default, on RHEL8 this value is 0.

The system needs to detect a hard lockup first to be able to panic. The detection is controlled by the nmi_watchdog parameter.

nmi_watchdog

Controls whether lockup detection mechanisms (watchdogs) are active or not. This parameter is of integer type.

ValueEffect

0

disables lockup detector

1

enables lockup detector

The hard lockup detector monitors each CPU for its ability to respond to interrupts.

watchdog_thresh

Controls frequency of watchdog hrtimer, NMI events, and soft/hard lockup thresholds.

Default thresholdSoft lockup threshold

10 seconds

2 * watchdog_thresh

Setting this parameter to zero disables lockup detection altogether.

8.3. Spurious soft lockups in virtualized environments

The soft lockup firing on physical hosts, as described in What is a soft lockup, usually represents a kernel or hardware bug. The same phenomenon happening on guest operating systems in virtualized environments may represent a false warning.

Heavy work-load on a host or high contention over some specific resource such as memory, usually causes a spurious soft lockup firing. This is because the host may schedule out the guest CPU for a period longer than 20 seconds. Then when the guest CPU is again scheduled to run on the host, it experiences a time jump which triggers due timers. The timers include also watchdog hrtimer, which can consequently report a soft lockup on the guest CPU.

Because a soft lockup in a virtualized environment may be spurious, you should not enable the kernel parameters that would cause a system panic when a soft lockup is reported on a guest CPU.

Important

To understand soft lockups in guests, it is essential to know that the host schedules the guest as a task, and the guest then schedules its own tasks.

Chapter 9. Adjusting kernel parameters for database servers

There are different sets of kernel parameters which can affect performance of specific database applications. To secure efficient operation of database servers and databases, configure the respective kernel parameters accordingly.

9.1. Introduction to database servers

A database server is a service that provides features of a database management system (DBMS). DBMS provides utilities for database administration and interacts with end users, applications, and databases.

Red Hat Enterprise Linux 9 provides the following database management systems:

  • MariaDB 10.5
  • MariaDB 10.11 - available since RHEL 9.4
  • MySQL 8.0
  • PostgreSQL 13
  • PostgreSQL 15 - available since RHEL 9.2
  • PostgreSQL 16 - available since RHEL 9.4
  • Redis 6

9.2. Parameters affecting performance of database applications

The following kernel parameters affect performance of database applications.

fs.aio-max-nr

Defines the maximum number of asynchronous I/O operations the system can handle on the server.

Note

Raising the fs.aio-max-nr parameter produces no additional changes beyond increasing the aio limit.

fs.file-max

Defines the maximum number of file handles (temporary file names or IDs assigned to open files) the system supports at any instance.

The kernel dynamically allocates file handles whenever a file handle is requested by an application. The kernel however does not free these file handles when they are released by the application. The kernel recycles these file handles instead. This means that over time the total number of allocated file handles will increase even though the number of currently used file handles may be low.

kernel.shmall
Defines the total number of shared memory pages that can be used system-wide. To use the entire main memory, the value of the kernel.shmall parameter should be ≤ total main memory size.
kernel.shmmax
Defines the maximum size in bytes of a single shared memory segment that a Linux process can allocate in its virtual address space.
kernel.shmmni
Defines the maximum number of shared memory segments the database server is able to handle.
net.ipv4.ip_local_port_range
Defines the port range the system can use for programs which want to connect to a database server without a specific port number.
net.core.rmem_default
Defines the default receive socket memory through Transmission Control Protocol (TCP).
net.core.rmem_max
Defines the maximum receive socket memory through Transmission Control Protocol (TCP).
net.core.wmem_default
Defines the default send socket memory through Transmission Control Protocol (TCP).
net.core.wmem_max
Defines the maximum send socket memory through Transmission Control Protocol (TCP).
vm.dirty_bytes / vm.dirty_ratio
Defines a threshold in bytes / in percentage of dirty-able memory at which a process generating dirty data is started in the write() function.
Note

Either vm.dirty_bytes or vm.dirty_ratio can be specified at a time.

vm.dirty_background_bytes / vm.dirty_background_ratio
Defines a threshold in bytes / in percentage of dirty-able memory at which the kernel tries to actively write dirty data to hard-disk.
Note

Either vm.dirty_background_bytes or vm.dirty_background_ratio can be specified at a time.

vm.dirty_writeback_centisecs

Defines a time interval between periodic wake-ups of the kernel threads responsible for writing dirty data to hard-disk.

This kernel parameters measures in 100th’s of a second.

vm.dirty_expire_centisecs

Defines the time after which dirty data is old enough to be written to hard-disk.

This kernel parameters measures in 100th’s of a second.

Chapter 10. Getting started with kernel logging

Log files are files that contain messages about the system, including the kernel, services, and applications running on it. The logging system in Red Hat Enterprise Linux is based on the built-in syslog protocol. Various utilities use this system to record events and organize them into log files. These files are useful when auditing the operating system or troubleshooting problems.

10.1. What is the kernel ring buffer

During the boot process, the console provides a lot of important information about the initial phase of the system startup. To avoid loss of the early messages the kernel utilizes what is called a ring buffer. This buffer stores all messages, including boot messages, generated by the printk() function within the kernel code. The messages from the kernel ring buffer are then read and stored in log files on permanent storage, for example, by the syslog service.

The buffer mentioned above is a cyclic data structure which has a fixed size, and is hard-coded into the kernel. Users can display data stored in the kernel ring buffer through the dmesg command or the /var/log/boot.log file. When the ring buffer is full, the new data overwrites the old.

Additional resources

  • syslog(2) and dmesg(1) manual page

10.2. Role of printk on log-levels and kernel logging

Each message the kernel reports has a log-level associated with it that defines the importance of the message. The kernel ring buffer, as described in What is the kernel ring buffer, collects kernel messages of all log-levels. It is the kernel.printk parameter that defines what messages from the buffer are printed to the console.

The log-level values break down in this order:

0
Kernel emergency. The system is unusable.
1
Kernel alert. Action must be taken immediately.
2
Condition of the kernel is considered critical.
3
General kernel error condition.
4
General kernel warning condition.
5
Kernel notice of a normal but significant condition.
6
Kernel informational message.
7
Kernel debug-level messages.

By default, kernel.printk in RHEL 9 contains the following four values:

# sysctl kernel.printk
kernel.printk = 7	4	1	7

The four values define the following, in order:

  1. Console log-level, defines the lowest priority of messages printed to the console.
  2. Default log-level for messages without an explicit log-level attached to them.
  3. Sets the lowest possible log-level configuration for the console log-level.
  4. Sets default value for the console log-level at boot time.

    Each of these values above defines a different rule for handling error messages.

Important

The default 7 4 1 7 printk value allows for better debugging of kernel activity. However, when coupled with a serial console, this printk setting might cause intense I/O bursts that might lead to a RHEL system becoming temporarily unresponsive. To avoid these situations, setting a printk value of 4 4 1 7 typically works, but at the expense of losing the extra debugging information.

Also note that certain kernel command line parameters, such as quiet or debug, change the default kernel.printk values.

Additional resources

  • syslog(2) manual page

Chapter 11. Reinstalling GRUB

You can reinstall the GRUB boot loader to fix certain problems, usually caused by an incorrect installation of GRUB, missing files, or a broken system. You can resolve these issues by restoring the missing files and updating the boot information.

Reasons to reinstall GRUB:

  • Upgrading the GRUB boot loader packages.
  • Adding the boot information to another drive.
  • The user requires the GRUB boot loader to control installed operating systems. However, some operating systems are installed with their own boot loaders and reinstalling GRUB returns control to the desired operating system.
Note

GRUB restores files only if they are not corrupted.

11.1. Reinstalling GRUB on BIOS-based machines

You can reinstall the GRUB boot loader on your BIOS-based system. Always reinstall GRUB after updating the GRUB packages.

Important

This overwrites the existing GRUB to install the new GRUB. Ensure that the system does not cause data corruption or boot crash during the installation.

Procedure

  1. Reinstall GRUB on the device where it is installed. For example, if sda is your device:

    # grub2-install /dev/sda
  2. Reboot your system for the changes to take effect:

    # reboot

Additional resources

  • The grub-install(1) man page

11.2. Reinstalling GRUB on UEFI-based machines

You can reinstall the GRUB boot loader on your UEFI-based system.

Important

Ensure that the system does not cause data corruption or boot crash during the installation.

Procedure

  1. Reinstall the grub2-efi and shim boot loader files:

    # yum reinstall grub2-efi shim
  2. Reboot your system for the changes to take effect:

    # reboot

11.3. Reinstalling GRUB on IBM Power machines

You can reinstall the GRUB boot loader on the Power PC Reference Platform (PReP) boot partition of your IBM Power system. Always reinstall GRUB after updating the GRUB packages.

Important

This overwrites the existing GRUB to install the new GRUB. Ensure that the system does not cause data corruption or boot crash during the installation.

Procedure

  1. Determine the disk partition that stores GRUB:

    # bootlist -m normal -o
    sda1
  2. Reinstall GRUB on the disk partition:

    # grub2-install partition

    Replace partition with the GRUB partition that you found in the previous step, such as /dev/sda1.

  3. Reboot your system for the changes to take effect:

    # reboot

Additional resources

  • The grub-install(1) man page

11.4. Resetting GRUB

Resetting GRUB completely removes all GRUB configuration files and system settings and reinstalls the bootloader. You can reset all the configuration settings to their default values, and thus fix failures caused by corrupted files and incorrect configuration.

Important

The following procedure will remove all the customization the user has made.

Procedure

  1. Remove the configuration files.

    # rm /etc/grub.d/*
    # rm /etc/sysconfig/grub
  2. Reinstall packages.

    • On BIOS-based machines, enter:

      # yum reinstall grub2-tools
    • On UEFI-based machines, enter:

      # yum reinstall grub2-efi shim grub2-tools
  3. Rebuild the grub.cfg file for the changes to take effect.

    • On BIOS-based machines, enter:

      # grub2-mkconfig -o /boot/grub2/grub.cfg
    • On UEFI-based machines, enter:

      # grub2-mkconfig -o /boot/efi/EFI/redhat/grub.cfg
  4. Follow Reinstalling GRUB procedure to restore GRUB on the /boot/ partition.

Chapter 12. Installing kdump

The kdump service is installed and activated by default on the new versions of Red Hat Enterprise Linux 9 installations. With the provided information and procedures, learn what kdump is and how to install if kdump is not enabled by default.

12.1. What is kdump

kdump is a service which provides a crash dumping mechanism and generates a dump file, known as crash dump or a vmcore file. The vmcore file has the contents of the system memory that helps in analysis and troubleshooting. kdump uses the kexec system call to boot into the second kernel, a capture kernel without a reboot and then captures the contents of the crashed kernel’s memory and saves it into a file. The second kernel is available in a reserved part of the system memory.

Important

A kernel crash dump can be the only information available if a system failure occur. Therefore, operational kdump is important in mission-critical environments. Red Hat advises to regularly update and test kexec-tools in your normal kernel update cycle. This is especially important when you install new kernel features.

You can enable kdump for all installed kernels on a machine or only for specified kernels. This is useful when there are multiple kernels used on a machine, some of which are stable enough that there is no concern that they could crash. When you install kdump, a default /etc/kdump.conf file is created. The /etc/kdump.conf file includes the default minimum kdump configuration, which you can edit to customize the kdump configuration.

12.2. Installing kdump using Anaconda

The Anaconda installer provides a graphical interface screen for kdump configuration during an interactive installation. The installer screen is titled as KDUMP and is available from the main Installation Summary screen. You can enable kdump and reserve the required amount of memory.

Procedure

  1. Under the KDUMP field, enable kdump if not already enabled.

    Enable kdump during RHEL installation
  2. Under Kdump Memory Reservation, select Manual` if you must customize the memory reserve.
  3. Under KDUMP field, in Memory To Be Reserved (MB), set the required memory reserve for kdump.

    Memory setting

12.3. Installing kdump on the command line

Some installation options, such as custom Kickstart installations, in some cases do not install or enable kdump by default. If this is your case, follow the procedure below.

Prerequisites

  • An active RHEL subscription.
  • A repository containing the kexec-tools package for your system CPU architecture.
  • Fulfilled requirements for kdump configurations and targets. For details, see Supported kdump configurations and targets.

Procedure

  1. Check whether kdump is installed on your system:

    # rpm -q kexec-tools

    Output if the package is installed:

    # kexec-tools-2.0.22-13.el9.x86_64

    Output if the package is not installed:

    package kexec-tools is not installed
  2. Install kdump and other necessary packages by:

    # dnf install kexec-tools

Chapter 13. Configuring kdump on the command line

The memory for kdump is reserved during the system boot. The memory size is configured in the system’s Grand Unified Bootloader (GRUB) configuration file. The memory size depends on the crashkernel= value specified in the configuration file and the size of the system’s physical memory.

13.1. Estimating the kdump size

When planning and building your kdump environment, it is important to know how much space the crash dump file requires.

The makedumpfile --mem-usage command estimates how much space the crash dump file requires. It generates a memory usage report. The report helps you determine the dump level and which pages are safe to be excluded.

Procedure

  • Execute the following command to generate a memory usage report:

    # makedumpfile --mem-usage /proc/kcore
    
    
    TYPE        PAGES    EXCLUDABLE    DESCRIPTION
    -------------------------------------------------------------
    ZERO          501635      yes        Pages filled with zero
    CACHE         51657       yes        Cache pages
    CACHE_PRIVATE 5442        yes        Cache pages + private
    USER          16301       yes        User process pages
    FREE          77738211    yes        Free pages
    KERN_DATA     1333192     no         Dumpable kernel data
Important

The makedumpfile --mem-usage command reports required memory in pages. This means that you must calculate the size of memory in use against the kernel page size.

By default the RHEL kernel uses 4 KB sized pages on AMD64 and Intel 64 CPU architectures, and 64 KB sized pages on IBM POWER architectures.

13.2. Configuring kdump memory usage on RHEL 9

The kexec-tools package maintains the default crashkernel= memory reservation values. The kdump service uses the default value to reserve the crash kernel memory for each kernel. The default value can also serve as the reference base value to estimate the required memory size when you set the crashkernel= value manually. The minimum size of the crash kernel can vary depending on the hardware and machine specifications.

The automatic memory allocation for kdump also varies based on the system hardware architecture and available memory size. For example, on AMD and Intel 64-bit architectures, the default value for the crashkernel= parameter will work only when the available memory is more than 1 GB. The kexec-tools utility configures the following default memory reserves on AMD64 and Intel 64-bit architecture:

crashkernel=1G-4G:192M,4G-64G:256M,64G:512M

You can also run kdumpctl estimate to query a rough estimate value without triggering a crash. The estimated crashkernel= value might not be an accurate one but can serve as a reference to set an appropriate crashkernel= value.

Note

The crashkernel=auto option in the boot command line is no longer supported on RHEL 9 and later releases.

Prerequisites

  • You have root permissions on the system.
  • You have fulfilled kdump requirements for configurations and targets. For details, see Supported kdump configurations and targets.
  • You have installed the zipl utility if it is the IBM Z system.

Procedure

  1. Configure the default value for crash kernel.

    # kdumpctl reset-crashkernel --kernel=ALL

    When configuring the crashkernel= value, test the configuration by rebooting with kdump enabled. If the kdump kernel fails to boot, increase the memory size gradually to set an acceptable value.

  2. To use a custom crashkernel= value:

    1. Configure the required memory reserve.

      crashkernel=192M

      Alternatively, you can set the amount of reserved memory to a variable depending on the total amount of installed memory using the syntax crashkernel=<range1>:<size1>,<range2>:<size2>. For example:

      crashkernel=1G-4G:192M,2G-64G:256M

      The example reserves 192 MB of memory if the total amount of system memory is 1 GB or higher and lower than 4 GB. If the total amount of memory is more than 4 GB, 256 MB is reserved for kdump.

    2. (Optional) Offset the reserved memory.

      Some systems require to reserve memory with a certain fixed offset since crashkernel reservation is very early, and it wants to reserve some area for special usage. If the offset is set, the reserved memory begins there. To offset the reserved memory, use the following syntax:

      crashkernel=192M@16M

      The example reserves 192 MB of memory starting at 16 MB (physical address 0x01000000). If you offset to 0 or do not specify a value, kdump offsets the reserved memory automatically. You can also offset memory when setting a variable memory reservation by specifying the offset as the last value. For example, crashkernel=1G-4G:192M,2G-64G:256M@16M.

    3. Update the bootloader configuration.

      # grubby --update-kernel ALL --args "crashkernel=<custom-value>"

      The <custom-value> must contain the custom crashkernel= value that you have configured for the crash kernel.

  3. Reboot for changes to take effect.

    # reboot

Verification

Cause the kernel to crash by activating the sysrq key. The address-YYYY-MM-DD-HH:MM:SS/vmcore file is saved to the target location as specified in the /etc/kdump.conf file. If you choose the default target location, the vmcore file is saved in the partition mounted under /var/crash/.

Warning

The commands to test kdump configuration will cause the kernel to crash with data loss. Follow the instructions with care and do not use an active production system to test the kdump configuration

  1. Activate the sysrq key to boot into the kdump kernel.

    # echo c > /proc/sysrq-trigger

    The command causes the kernel to crash and reboots the kernel if required.

  2. Display the /etc/kdump.conf file and check if the vmcore file is saved in the target destination.

13.3. Configuring the kdump target

The crash dump is usually stored as a file in a local file system, written directly to a device. Alternatively, you can set up for the crash dump to be sent over a network using the NFS or SSH protocols. Only one of these options to preserve a crash dump file can be set at a time. The default behavior is to store it in the /var/crash/ directory of the local file system.

Prerequisites

Procedure

  • To store the crash dump file in /var/crash/ directory of the local file system, edit the /etc/kdump.conf file and specify the path:

    path /var/crash

    The option path /var/crash represents the path to the file system in which kdump saves the crash dump file.

    Note
    • When you specify a dump target in the /etc/kdump.conf file, then the path is relative to the specified dump target.
    • When you do not specify a dump target in the /etc/kdump.conf file, then the path represents the absolute path from the root directory.

    Depending on what is mounted in the current system, the dump target and the adjusted dump path are taken automatically.

    To secure the crash dump file and the accompanying files produced by kdump, you should set up proper attributes for the target destination directory, such as user permissions and SELinux contexts. Additionally, you can define a script, for example kdump_post.sh in the kdump.conf file as follows:

    kdump_post <path_to_kdump_post.sh>

    The kdump_post directive specifies a shell script or a command that is executed after kdump has completed capturing and saving a crash dump to the specified destination. You can use this mechanism to extend the functionality of kdump to perform actions including the adjustment of file permissions.

Example 13.1. The kdump target configuration

# grep -v ^# /etc/kdump.conf | grep -v ^$
ext4 /dev/mapper/vg00-varcrashvol
path /var/crash
core_collector makedumpfile -c --message-level 1 -d 31

Here, the dump target is specified (ext4 /dev/mapper/vg00-varcrashvol), and thus mounted at /var/crash. The path option is also set to /var/crash, so the kdump saves the vmcore file in the /var/crash/var/crash directory.

  • To change the local directory in which the crash dump is to be saved, as root, edit the /etc/kdump.conf configuration file:

    1. Remove the hash sign (#) from the beginning of the #path /var/crash line.
    2. Replace the value with the intended directory path. For example:

      path /usr/local/cores
      Important

      In Red Hat Enterprise Linux 9, the directory defined as the kdump target using the path directive must exist when the kdump systemd service starts to avoid failures. This behavior is different from versions earlier than RHEL, where the directory is created automatically if it did not exist when the service starts.

  • To write the file to a different partition, edit the /etc/kdump.conf configuration file:

    1. Remove the hash sign (#) from the beginning of the #ext4 line, depending on your choice.

      • device name (the #ext4 /dev/vg/lv_kdump line)
      • file system label (the #ext4 LABEL=/boot line)
      • UUID (the #ext4 UUID=03138356-5e61-4ab3-b58e-27507ac41937 line)
    2. Change the file system type and the device name, label or UUID, to the required values. The correct syntax for specifying UUID values is both UUID="correct-uuid" and UUID=correct-uuid. For example:

      ext4 UUID=03138356-5e61-4ab3-b58e-27507ac41937
      Important

      It is recommended to specify storage devices using a LABEL= or UUID=. Disk device names such as /dev/sda3 are not guaranteed to be consistent across reboot.

      When you use Direct Access Storage Device (DASD) on IBM Z hardware, ensure the dump devices are correctly specified in /etc/dasd.conf before you proceed with kdump.

  • To write the crash dump directly to a device, edit the /etc/kdump.conf configuration file:

    1. Remove the hash sign (#) from the beginning of the #raw /dev/vg/lv_kdump line.
    2. Replace the value with the intended device name. For example:

      raw /dev/sdb1
  • To store the crash dump to a remote machine using the NFS protocol:

    1. Remove the hash sign (#) from the beginning of the #nfs my.server.com:/export/tmp line.
    2. Replace the value with a valid hostname and directory path. For example:

      nfs penguin.example.com:/export/cores
    3. Restart the kdump service for the changes to take effect:

      sudo systemctl restart kdump.service
      Note

      When using the NFS directive to specify the NFS target, kdump.service automatically attempts to mount the NFS target to check the disk space. There is no need to mount the NFS target beforehand. To prevent kdump.service from mounting the target, use the dracut_args --mount directive in kdump.conf so that kdump.service calls the dracut utility with the --mount argument to specify the NFS target.

  • To store the crash dump to a remote machine using the SSH protocol:

    1. Remove the hash sign (#) from the beginning of the #ssh user@my.server.com line.
    2. Replace the value with a valid username and hostname.
    3. Include your SSH key in the configuration.

      1. Remove the hash sign from the beginning of the #sshkey /root/.ssh/kdump_id_rsa line.
      2. Change the value to the location of a key valid on the server you are trying to dump to. For example:

        ssh john@penguin.example.com
        sshkey /root/.ssh/mykey

13.4. Configuring the kdump core collector

The kdump service uses a core_collector program to capture the crash dump image. In RHEL, the makedumpfile utility is the default core collector. It helps shrink the dump file by:

  • Compressing the size of a crash dump file and copying only necessary pages using various dump levels.
  • Excluding unnecessary crash dump pages.
  • Filtering the page types to be included in the crash dump.

Syntax

core_collector makedumpfile -l --message-level 1 -d 31

Options

  • -c, -l or -p: specify compress dump file format by each page using either, zlib for -c option, lzo for -l option or snappy for -p option.
  • -d (dump_level): excludes pages so that they are not copied to the dump file.
  • --message-level : specify the message types. You can restrict outputs printed by specifying message_level with this option. For example, specifying 7 as message_level prints common messages and error messages. The maximum value of message_level is 31

Prerequisites

Procedure

  1. As root, edit the /etc/kdump.conf configuration file and remove the hash sign ("#") from the beginning of the #core_collector makedumpfile -l --message-level 1 -d 31.
  2. To enable crash dump file compression, execute:
core_collector makedumpfile -l --message-level 1 -d 31

The -l option specifies the dump compressed file format. The -d option specifies dump level as 31. The --message-level option specifies message level as 1.

Also, consider following examples with the -c and -p options:

  • To compress a crash dump file using -c:
core_collector makedumpfile -c -d 31 --message-level 1
  • To compress a crash dump file using -p:
core_collector makedumpfile -p -d 31 --message-level 1

Additional resources

13.5. Configuring the kdump default failure responses

By default, when kdump fails to create a crash dump file at the configured target location, the system reboots and the dump is lost in the process. You can change the default failure response and configure kdump to perform a different operation in case it fails to save the core dump to the primary target. The additional actions are:

dump_to_rootfs
Saves the core dump to the root file system.
reboot
Reboots the system, losing the core dump in the process.
halt
Stops the system, losing the core dump in the process.
poweroff
Power the system off, losing the core dump in the process.
shell
Runs a shell session from within the initramfs, you can record the core dump manually.
final_action
Enables additional operations such as reboot, halt, and poweroff after a successful kdump or when shell or dump_to_rootfs failure action completes. The default is reboot.
failure_action
Specifies the action to perform when a dump might fail in a kernel crash. The default is reboot.

Prerequisites

Procedure

  1. As root, remove the hash sign (#) from the beginning of the #failure_action line in the /etc/kdump.conf configuration file.
  2. Replace the value with a desired action.

    failure_action poweroff

Additional resources

13.6. Configuration file for kdump

The configuration file for kdump kernel is /etc/sysconfig/kdump. This file controls the kdump kernel command line parameters. For most configurations, use the default options. However, in some scenarios you might need to modify certain parameters to control the kdump kernel behavior. For example, modifying the KDUMP_COMMANDLINE_APPEND option to append the kdump kernel command-line to obtain a detailed debugging output or the KDUMP_COMMANDLINE_REMOVE option to remove arguments from the kdump command line.

KDUMP_COMMANDLINE_REMOVE

This option removes arguments from the current kdump command line. It removes parameters that may cause kdump errors or kdump kernel boot failures. These parameters may have been parsed from the previous KDUMP_COMMANDLINE process or inherited from the /proc/cmdline file.

When this variable is not configured, it inherits all values from the /proc/cmdline file. Configuring this option also provides information that is helpful in debugging an issue.

To remove certain arguments, add them to KDUMP_COMMANDLINE_REMOVE as follows:

KDUMP_COMMANDLINE_REMOVE="hugepages hugepagesz slub_debug quiet log_buf_len swiotlb"
KDUMP_COMMANDLINE_APPEND

This option appends arguments to the current command line. These arguments may have been parsed by the previous KDUMP_COMMANDLINE_REMOVE variable.

For the kdump kernel, disabling certain modules such as mce, cgroup, numa, hest_disable can help prevent kernel errors. These modules may consume a significant portion of the kernel memory reserved for kdump or cause kdump kernel boot failures.

To disable memory cgroups on the kdump kernel command line, run the command as follows:

KDUMP_COMMANDLINE_APPEND="cgroup_disable=memory"

Additional resources

  • The Documentation/admin-guide/kernel-parameters.txt file
  • The /etc/sysconfig/kdump file

13.7. Testing the kdump configuration

After configuring kdump, you must manually test a system crash and ensure that the vmcore file is generated in the defined kdump target. The vmcore file is captured from the context of the freshly booted kernel and therefore has critical information to help debug a kernel crash.

Warning

Do not test kdump on active production systems. The commands to test kdump will cause the kernel to crash with loss of data. Depending on your system architecture, ensure that you schedule significant maintenance time because kdump testing might require several reboots with a long boot time.

If the vmcore file is not generated during the kdump test, identify and fix issues before you run the test again for a successful kdump testing.

Important

Ensure that you schedule significant maintenance time, because kdump testing might require several reboots with a long boot time.

If you make any manual system modifications, you must test the kdump configuration at the end of any system modification. For example, if you make any of the following changes, ensure that you test the kdump configuration for an optimal kdump performance:

  • Package upgrades.
  • Hardware level changes, for example, storage or networking changes.
  • Firmware and BIOS upgrades.
  • New installation and application upgrades that include third party modules.
  • If you use the hot-plugging mechanism to add more memory on hardware that support this mechanism.
  • After you make changes in the /etc/kdump.conf or /etc/sysconfig/kdump file.

Prerequisites

  • You have root permissions on the system.
  • You have saved all important data. The commands to test kdump cause the kernel to crash with loss of data.
  • You have scheduled significant machine maintenance time depending on the system architecture.

Procedure

  1. Enable the kdump service:

    # kdumpctl restart
  2. Check the status of the kdump service. With the kdumpctl command, you can print the output at the console.

    # kdumpctl status
      kdump:Kdump is operational

    Alternatively, if you use the systemctl command, the output prints in the systemd journal.

  3. Initiate a kernel crash to test the kdump configuration. The sysrq-trigger key combination causes the kernel to crash and might reboot the system if required.

    # echo c > /proc/sysrq-trigger

    On a kernel reboot, the address-YYYY-MM-DD-HH:MM:SS/vmcore file is created at the location you have specified in the /etc/kdump.conf file. The default is /var/crash/.

Additional resources

13.8. Files produced by kdump after system crash

After your system crashes, the kdump service captures the kernel memory in a dump file (vmcore) and it also generates additional diagnostic files to aid in troubleshooting and post-mortem analysis.

Files produced by kdump:

  • vmcore - main kernel memory dump file containing system memory at the time of the crash. It includes data as per the configuration of the core_collector program specified in kdump configuration. By default the kernel data structures, process information, stack traces, and other diagnostic information.
  • vmcore-dmesg.txt - contents of the kernel ring buffer log (dmesg) from the primary kernel that panicked.
  • kexec-dmesg.log - contains kernel and system log messages from the execution of the secondary kexec kernel that collects the vmcore data.

13.9. Enabling and disabling the kdump service

You can configure to enable or disable the kdump functionality on a specific kernel or on all installed kernels. You must routinely test the kdump functionality and validate that it is working properly.

Prerequisites

  • You have root permissions on the system.
  • You have completed kdump requirements for configurations and targets. See Supported kdump configurations and targets.
  • All configurations for installing kdump are set up as required.

Procedure

  • Enable the kdump service for multi-user.target:

    # systemctl enable kdump.service
  • Start the service in the current session:

    # systemctl start kdump.service
  • Stop the kdump service:

    # systemctl stop kdump.service
  • Disable the kdump service:

    # systemctl disable kdump.service
Warning

It is recommended to set kptr_restrict=1 as default. When kptr_restrict is set to (1) as default, the kdumpctl service loads the crash kernel even if Kernel Address Space Layout (KASLR) is enabled or not enabled.

If kptr_restrict is not set to 1 and KASLR is enabled, the contents of /proc/kore file are generated as all zeros. The kdumpctl service fails to access the /proc/kcore file and load the crash kernel. The kexec-kdump-howto.txt file displays a warning message, which recommends you to set kptr_restrict=1. Verify for the following in the sysctl.conf file to ensure that kdumpctl service loads the crash kernel:

  • Kernel kptr_restrict=1 in the sysctl.conf file.

13.10. Preventing kernel drivers from loading for kdump

You can control the capture kernel from loading certain kernel drivers by adding the KDUMP_COMMANDLINE_APPEND= variable in the /etc/sysconfig/kdump configuration file. By using this method, you can prevent the kdump initial RAM disk image initramfs from loading the specified kernel module. This helps to prevent the out-of-memory (OOM) killer errors or other crash kernel failures.

You can append the KDUMP_COMMANDLINE_APPEND= variable using one of the following configuration options:

  • rd.driver.blacklist=<modules>
  • modprobe.blacklist=<modules>

Prerequisites

  • You have root permissions on the system.

Procedure

  1. Display the list of modules that are loaded to the currently running kernel. Select the kernel module that you intend to block from loading.

    $ lsmod
    
    Module                  Size  Used by
    fuse                  126976  3
    xt_CHECKSUM            16384  1
    ipt_MASQUERADE         16384  1
    uinput                 20480  1
    xt_conntrack           16384  1
  2. Update the KDUMP_COMMANDLINE_APPEND= variable in the /etc/sysconfig/kdump file. For example:

    KDUMP_COMMANDLINE_APPEND="rd.driver.blacklist=hv_vmbus,hv_storvsc,hv_utils,hv_netvsc,hid-hyperv"

    Also, consider the following example using the modprobe.blacklist=<modules> configuration option:

    KDUMP_COMMANDLINE_APPEND="modprobe.blacklist=emcp modprobe.blacklist=bnx2fc modprobe.blacklist=libfcoe modprobe.blacklist=fcoe"
  3. Restart the kdump service:

    # systemctl restart kdump

Additional resources

  • dracut.cmdline man page

13.11. Running kdump on systems with encrypted disk

When you run a LUKS encrypted partition, systems require certain amount of available memory. If the system has less than the required amount of available memory, the cryptsetup utility fails to mount the partition. As a result, capturing the vmcore file to an encrypted target location fails in the second kernel (capture kernel).

The kdumpctl estimate command helps you estimate the amount of memory you need for kdump. kdumpctl estimate prints the recommended crashkernel value, which is the most suitable memory size required for kdump.

The recommended crashkernel value is calculated based on the current kernel size, kernel module, initramfs, and the LUKS encrypted target memory requirement.

In case you are using the custom crashkernel= option, kdumpctl estimate prints the LUKS required size value. The value is the memory size required for LUKS encrypted target.

Procedure

  1. Print the estimate crashkernel= value:

    # *kdumpctl estimate*
    
    Encrypted kdump target requires extra memory, assuming using the keyslot with minimum memory requirement
       Reserved crashkernel:    256M
       Recommended crashkernel: 652M
    
       Kernel image size:   47M
       Kernel modules size: 8M
       Initramfs size:      20M
       Runtime reservation: 64M
       LUKS required size:  512M
       Large modules: <none>
       WARNING: Current crashkernel size is lower than recommended size 652M.
  2. Configure the amount of required memory by increasing the crashkernel= value.
  3. Reboot the system.
Note

If the kdump service still fails to save the dump file to the encrypted target, increase the crashkernel= value as required.

Chapter 14. Enabling kdump

For your Red Hat Enterprise Linux 9 systems, you can configure to enable or disable the kdump functionality on a specific kernel or on all installed kernels. However, you must routinely test the kdump functionality and validate that it’s working properly.

14.1. Enabling kdump for all installed kernels

The kdump service starts by enabling kdump.service after the kexec tool is installed. You can enable and start the kdump service for all kernels installed on the machine.

Prerequisites

  • You have administrator privileges.

Procedure

  1. Add the crashkernel= command-line parameter to all installed kernels.

    # grubby --update-kernel=ALL --args="crashkernel=xxM"

    xxM is the required memory in megabytes.

  2. Enable the kdump service.

    # systemctl enable --now kdump.service

Verification

  • Check that the kdump service is running.

    # systemctl status kdump.service
    
    ○ kdump.service - Crash recovery kernel arming
         Loaded: loaded (/usr/lib/systemd/system/kdump.service; enabled; vendor preset: disabled)
         Active: active (live)

14.2. Enabling kdump for a specific installed kernel

You can enable the kdump service for a specific kernel on the machine.

Prerequisites

  • You have administrator privileges.

Procedure

  1. List the kernels installed on the machine.

    # ls -a /boot/vmlinuz-*
    /boot/vmlinuz-0-rescue-2930657cd0dc43c2b75db480e5e5b4a9
    /boot/vmlinuz-4.18.0-330.el8.x86_64
    /boot/vmlinuz-4.18.0-330.rt7.111.el8.x86_64
  2. Add a specific kdump kernel to the system’s Grand Unified Bootloader (GRUB) configuration.

    For example:

    # grubby --update-kernel=vmlinuz-4.18.0-330.el8.x86_64 --args="crashkernel=xxM"

    xxM is the required memory reserve in megabytes.

  3. Enable the kdump service.

    # systemctl enable --now kdump.service

Verification

  • Check that the kdump service is running.

    # systemctl status kdump.service
    
    ○ kdump.service - Crash recovery kernel arming
         Loaded: loaded (/usr/lib/systemd/system/kdump.service; enabled; vendor preset: disabled)
         Active: active (live)

14.3. Disabling the kdump service

You can stop the kdump.service and disable the service from starting on your Red Hat Enterprise Linux 9 systems.

Prerequisites

Procedure

  1. To stop the kdump service in the current session:

    # systemctl stop kdump.service
  2. To disable the kdump service:

    # systemctl disable kdump.service
Warning

It is recommended to set kptr_restrict=1 as default. When kptr_restrict is set to (1) as default, the kdumpctl service loads the crash kernel even if Kernel Address Space Layout (KASLR) is enabled or not enabled.

If kptr_restrict is not set to 1 and KASLR is enabled, the contents of /proc/kore file are generated as all zeros. The kdumpctl service fails to access the /proc/kcore file and load the crash kernel. The kexec-kdump-howto.txt file displays a warning message, which recommends you to set kptr_restrict=1. Verify for the following in the sysctl.conf file to ensure that kdumpctl service loads the crash kernel:

  • Kernel kptr_restrict=1 in the sysctl.conf file.

Additional resources

Chapter 15. Supported kdump configurations and targets

The kdump mechanism is a feature of the Linux kernel that generates a crash dump file when a kernel crash occurs. The kernel dump file has critical information that helps to analyze and determine the root cause of a kernel crash. The crash can be because of various factors, hardware issues or third-party kernel modules problems, to name a few.

With the provided information and procedures, you understand the supported configurations and targets on your Red Hat Enterprise Linux 9 systems and properly configure kdump and validate it’s working.

15.1. Memory requirements for kdump

For kdump to capture a kernel crash dump and save it for further analysis, a part of the system memory should be permanently reserved for the capture kernel. When reserved, this part of the system memory is not available to the main kernel.

The memory requirements vary based on certain system parameters. One of the major factors is the system’s hardware architecture. To find out the exact machine architecture (such as Intel 64 and AMD64, also known as x86_64) and print it to standard output, use the following command:

$ uname -m

With the stated list of minimum memory requirements, you can set the appropriate memory size to automatically reserve a memory for kdump on the latest available versions. The memory size depends on the system’s architecture and total available physical memory.

Table 15.1. Minimum amount of reserved memory required for kdump

ArchitectureAvailable MemoryMinimum Reserved Memory

AMD64 and Intel 64 (x86_64)

1 GB to 4 GB

192 MB of RAM

4 GB to 64 GB

256 MB of RAM

64 GB and more

512 MB of RAM

64-bit ARM (4k pages)

1 GB to 4 GB

256 MB of RAM

4 GB to 64 GB

320 MB of RAM

64 GB and more

576 MB of RAM

64-bit ARM (64k pages)

1 GB to 4 GB

356 MB of RAM

4 GB to 64 GB

420 MB of RAM

64 GB and more

676 MB of RAM

IBM Power Systems (ppc64le)

2 GB to 4 GB

384 MB of RAM

4 GB to 16 GB

512 MB of RAM

16 GB to 64 GB

1 GB of RAM

64 GB to 128 GB

2 GB of RAM

128 GB and more

4 GB of RAM

IBM Z (s390x)

1 GB to 4 GB

192 MB of RAM

4 GB to 64 GB

256 MB of RAM

64 GB and more

512 MB of RAM

On many systems, kdump is able to estimate the amount of required memory and reserve it automatically. This behavior is enabled by default, but only works on systems that have more than a certain amount of total available memory, which varies based on the system architecture.

Important

The automatic configuration of reserved memory based on the total amount of memory in the system is a best effort estimation. The actual required memory may vary due to other factors such as I/O devices. Using not enough of memory might cause that a debug kernel is not able to boot as a capture kernel in case of a kernel panic. To avoid this problem, sufficiently increase the crash kernel memory.

15.2. Minimum threshold for automatic memory reservation

The kexec-tools utility, by default, configures the crashkernel command line parameter and reserves a certain amount of memory for kdump. On some systems however, it is still possible to allocate memory for kdump either by using the crashkernel=auto parameter in the boot loader configuration file, or by enabling this option in the graphical configuration utility. For this automatic reservation to work, a certain amount of total memory needs to be available in the system. The memory requirement varies based on the system’s architecture. If the system memory is less than the specified threshold value, you must configure the memory manually.

Table 15.2. Minimum amount of memory required for memory reservation

ArchitectureRequired Memory

AMD64 and Intel 64 (x86_64)

1 GB

IBM Power Systems (ppc64le)

2 GB

IBM  Z (s390x)

1 GB

64-bit ARM

1 GB

Note

The crashkernel=auto option in the boot command line is no longer supported on RHEL 9 and later releases.

15.3. Supported kdump targets

When a kernel crash occurs, the operating system saves the dump file on the configured or default target location. You can save the dump file either directly to a device, store as a file on a local file system, or send the dump file over a network. With the following list of dump targets, you can know the targets that are currently supported or not supported by kdump.

Table 15.3. kdump targets on RHEL 9

Target typeSupported TargetsUnsupported Targets

Physical Storage

  • Logical Volume Manager (LVM).
  • Thin provisioning volume.
  • Fibre Channel (FC) disks such as qla2xxx, lpfc, bnx2fc, and bfa.
  • An iSCSI software-configured logical device on a networked storage server.
  • The mdraid subsystem as a software RAID solution.
  • Hardware RAID such as smartpqi, hpsa, megaraid, mpt3sas, aacraid, and mpi3mr.
  • SCSI and SATA disks.
  • iSCSI and HBA offloads.
  • Hardware FCoE such as qla2xxx and lpfc.
  • Software FCoE such as bnx2fc. For software FCoE to function, additional memory configuration might be required.
  • BIOS RAID.
  • Software iSCSI with iBFT. Currently supported transports are bnx2i, cxgb3i, and cxgb4i.
  • Software iSCSI with hybrid device driver such as be2iscsi.
  • Fibre Channel over Ethernet (FCoE).
  • Legacy IDE.
  • GlusterFS servers.
  • GFS2 file system.
  • Clustered Logical Volume Manager (CLVM).
  • High availability LVM volumes (HA-LVM).

Network

  • Hardware using kernel modules such as igb, ixgbe, ice, i40e, e1000e, igc, tg3, bnx2x, bnxt_en, qede, cxgb4, be2net, enic, sfc, mlx4_en, mlx5_core, r8169, atlantic, nfp, and nicvf on 64-bit ARM architecture only.
  • Hardware using kernel modules such as sfc SRIOV, cxgb4vf, and pch_gbe.
  • IPv6 protocol.
  • Wireless connections.
  • InfiniBand networks.
  • VLAN network over bridge and team.

Hypervisor

  • Kernel-based virtual machines (KVM).
  • Xen hypervisor in certain configurations only.
  • ESXi 6.6, 6.7, 7.0.
  • Hyper-V 2012 R2 on RHEL Gen1 UP Guest only and later version.
 

Filesystem

The ext[234]fs, XFS, virtiofs, and NFS file systems.

The Btrfs file system.

Firmware

  • BIOS-based systems.
  • UEFI Secure Boot.
 

Additional resources

15.4. Supported kdump filtering levels

To reduce the size of the dump file, kdump uses the makedumpfile core collector to compress the data and also exclude unwanted information, for example, you can remove hugepages and hugetlbfs pages by using the -8 level. The levels that makedumpfile currently supports can be seen in the table for Filtering levels for `kdump` .

Table 15.4. Filtering levels for kdump

OptionDescription

1

Zero pages

2

Cache pages

4

Cache private

8

User pages

16

Free pages

Additional resources

15.5. Supported default failure responses

By default, when kdump fails to create a core dump, the operating system reboots. You can, however, configure kdump to perform a different operation in case it fails to save the core dump to the primary target.

dump_to_rootfs
Attempt to save the core dump to the root file system. This option is especially useful in combination with a network target: if the network target is unreachable, this option configures kdump to save the core dump locally. The system is rebooted afterwards.
reboot
Reboot the system, losing the core dump in the process.
halt
Halt the system, losing the core dump in the process.
poweroff
Power off the system, losing the core dump in the process.
shell
Run a shell session from within the initramfs, allowing the user to record the core dump manually.
final_action
Enable additional operations such as reboot, halt, and poweroff actions after a successful kdump or when shell or dump_to_rootfs failure action completes. The default final_action option is reboot.
failure_action
Specifies the action to perform when a dump might fail in the event of a kernel crash. The default failure_action option is reboot.

15.6. Using final_action parameter

When kdump succeeds or if kdump fails to save the vmcore file at the configured target, you can perform additional operations like reboot, halt, and poweroff by using the final_action parameter. If the final_action parameter is not specified, reboot is the default response.

Procedure

  1. To configure final_action, edit the /etc/kdump.conf file and add one of the following options:

    • final_action reboot
    • final_action halt
    • final_action poweroff
  2. Restart the kdump service for the changes to take effect.

    # kdumpctl restart

15.7. Using failure_action parameter

The failure_action parameter specifies the action to perform when a dump fails in the event of a kernel crash. The default action for failure_action is reboot, which reboots the system.

The parameter recognizes the following actions to take:

reboot
Reboots the system after a dump failure.
dump_to_rootfs
Saves the dump file on a root file system when a non-root dump target is configured.
halt
Halts the system.
poweroff
Stops the running operations on the system.
shell
Starts a shell session inside initramfs, from which you can manually perform additional recovery actions.

Procedure:

  1. To configure an action to take if the dump fails, edit the /etc/kdump.conf file and specify one of the failure_action options:

    • failure_action reboot
    • failure_action halt
    • failure_action poweroff
    • failure_action shell
    • failure_action dump_to_rootfs
  2. Restart the kdump service for the changes to take effect.

    # kdumpctl restart

Chapter 16. Firmware assisted dump mechanisms

Firmware assisted dump (fadump) is a dump capturing mechanism, provided as an alternative to the kdump mechanism on IBM POWER systems. The kexec and kdump mechanisms are useful for capturing core dumps on AMD64 and Intel 64 systems. However, some hardware such as mini systems and mainframe computers, leverage the onboard firmware to isolate regions of memory and prevent any accidental overwriting of data that is important to the crash analysis. The fadump utility, is optimized for the fadump mechanisms and their integration with RHEL on IBM POWER systems.

16.1. Firmware assisted dump on IBM PowerPC hardware

The fadump utility captures the vmcore file from a fully-reset system with PCI and I/O devices. This mechanism uses firmware to preserve memory regions during a crash and then reuses the kdump userspace scripts to save the vmcore file. The memory regions consist of all system memory contents, except the boot memory, system registers, and hardware Page Table Entries (PTEs).

The fadump mechanism offers improved reliability over the traditional dump type, by rebooting the partition and using a new kernel to dump the data from the previous kernel crash. The fadump requires an IBM POWER6 processor-based or later version hardware platform.

For further details about the fadump mechanism, including PowerPC specific methods of resetting hardware, see the /usr/share/doc/kexec-tools/fadump-howto.txt file.

Note

The area of memory that is not preserved, known as boot memory, is the amount of RAM required to successfully boot the kernel after a crash event. By default, the boot memory size is 256MB or 5% of total system RAM, whichever is larger.

Unlike kexec-initiated event, the fadump mechanism uses the production kernel to recover a crash dump. When booting after a crash, PowerPC hardware makes the device node /proc/device-tree/rtas/ibm.kernel-dump available to the proc filesystem (procfs). The fadump-aware kdump scripts, check for the stored vmcore, and then complete the system reboot cleanly.

16.2. Enabling firmware assisted dump mechanism

You can enhance the crash dumping capabilities of IBM POWER systems by enabling the firmware assisted dump (fadump) mechanism.

In the Secure Boot environment, the GRUB boot loader allocates a boot memory region, known as the Real Mode Area (RMA). The RMA has a size of 512 MB, which is divided among the boot components and, if a component exceeds its size allocation, GRUB fails with an out-of-memory (OOM) error.

Warning

Do not enable firmware assisted dump (fadump) mechanism in the Secure Boot environment on RHEL 9.1 and earlier versions. The GRUB boot loader fails with the following error:

error: ../../grub-core/kern/mm.c:376:out of memory.
Press any key to continue…

The system is recoverable only if you increase the default initramfs size due to the fadump configuration.

For information about workaround methods to recover the system, see the System boot ends in GRUB Out of Memory (OOM) article.

Prerequisites

  • You have root permissions on the system.

Procedure

  1. Install the kexec-tools package.
  2. Configure the default value for crashkernel.

    # kdumpctl reset-crashkernel --fadump=on --kernel=ALL
  3. (Optional) Reserve boot memory instead of the default value.

    # grubby --update-kernel ALL --args="fadump=on crashkernel=xxM"

    xxM is the required memory size in megabytes.

    Note

    When specifying boot configuration options, test the configurations by rebooting the kernel with kdump enabled. If the kdump kernel fails to boot, increase the crashkernel value gradually to set an appropriate value.

  4. Reboot for changes to take effect.

    # reboot

16.3. Firmware assisted dump mechanisms on IBM Z hardware

IBM Z systems support the following firmware assisted dump mechanisms:

  • Stand-alone dump (sadump)
  • VMDUMP

The kdump infrastructure is supported and utilized on IBM Z systems. However, using one of the firmware assisted dump (fadump) methods for IBM Z can provide various benefits:

  • The sadump mechanism is initiated and controlled from the system console, and is stored on an IPL bootable device.
  • The VMDUMP mechanism is similar to sadump. This tool is also initiated from the system console, but retrieves the resulting dump from hardware and copies it to the system for analysis.
  • These methods (similarly to other hardware based dump mechanisms) have the ability to capture the state of a machine in the early boot phase, before the kdump service starts.
  • Although VMDUMP contains a mechanism to receive the dump file into a Red Hat Enterprise Linux system, the configuration and control of VMDUMP is managed from the IBM Z Hardware console.

16.4. Using sadump on Fujitsu PRIMEQUEST systems

The Fujitsu sadump mechanism is designed to provide a fallback dump capture in an event when kdump is unable to complete successfully. The sadump mechanism is invoked manually from the system Management Board (MMB) interface. Using MMB, configure kdump like for an Intel 64 or AMD 64 server and then proceed to enable sadump.

Procedure

  1. Add or edit the following lines in the /etc/sysctl.conf file to ensure that kdump starts as expected for sadump:

    kernel.panic=0
    kernel.unknown_nmi_panic=1
    Warning

    In particular, ensure that after kdump, the system does not reboot. If the system reboots after kdump has failed to save the vmcore file, then it is not possible to invoke the sadump.

  2. Set the failure_action parameter in /etc/kdump.conf appropriately as halt or shell.

    failure_action shell

Additional resources

  • The FUJITSU Server PRIMEQUEST 2000 Series Installation Manual

Chapter 17. Analyzing a core dump

To determine the cause of the system crash, you can use the crash utility, which provides an interactive prompt very similar to the GNU Debugger (GDB). This utility allows you to interactively analyze a core dump created by kdump, netdump, diskdump or xendump as well as a running Linux system. Alternatively, you have the option to use Kernel Oops Analyzer or the Kdump Helper tool.

17.1. Installing the crash utility

With the provided information, understand the required packages and the procedure to install the crash utility. The crash utility may not be installed by default on your Red Hat Enterprise Linux 9 systems. crash is a tool to interactively analyze a system' state while it is running or after a kernel crash occurs and a core dump file is created. The core dump file is also known as the vmcore file.

Procedure

  1. Enable the relevant repositories:

    # subscription-manager repos --enable baseos repository
    # subscription-manager repos --enable appstream repository
    # subscription-manager repos --enable rhel-9-for-x86_64-baseos-debug-rpms
  2. Install the crash package:

    # dnf install crash
  3. Install the kernel-debuginfo package:

    # dnf install kernel-debuginfo

    The package kernel-debuginfo will correspond to the running kernel and provides the data necessary for the dump analysis.

17.2. Running and exiting the crash utility

With the provided information, understand the required parameters and the procedure to run and exit the crash utility. crash is a tool to interactively analyze a system' state while it is running or after a kernel crash occurs and a core dump file is created. The core dump file is also known as the vmcore file.

Prerequisites

  • Identify the currently running kernel (for example 5.14.0-1.el9.x86_64).

Procedure

  1. To start the crash utility, two necessary parameters need to be passed to the command:

    • The debug-info (a decompressed vmlinuz image), for example /usr/lib/debug/lib/modules/5.14.0-1.el9.x86_64/vmlinux provided through a specific kernel-debuginfo package.
    • The actual vmcore file, for example /var/crash/127.0.0.1-2021-09-13-14:05:33/vmcore

      The resulting crash command then looks like this:

      # crash /usr/lib/debug/lib/modules/5.14.0-1.el9.x86_64/vmlinux /var/crash/127.0.0.1-2021-09-13-14:05:33/vmcore

      Use the same <kernel> version that was captured by kdump.

      Example 17.1. Running the crash utility

      The following example shows analyzing a core dump created on September 13 2021 at 14:05 PM, using the 5.14.0-1.el9.x86_64 kernel.

      ...
      WARNING: kernel relocated [202MB]: patching 90160 gdb minimal_symbol values
      
            KERNEL: /usr/lib/debug/lib/modules/5.14.0-1.el9.x86_64/vmlinux
          DUMPFILE: /var/crash/127.0.0.1-2021-09-13-14:05:33/vmcore  [PARTIAL DUMP]
              CPUS: 2
              DATE: Mon Sep 13 14:05:16 2021
            UPTIME: 01:03:57
      LOAD AVERAGE: 0.00, 0.00, 0.00
             TASKS: 586
          NODENAME: localhost.localdomain
           RELEASE: 5.14.0-1.el9.x86_64
           VERSION: #1 SMP Wed Aug 29 11:51:55 UTC 2018
           MACHINE: x86_64  (2904 Mhz)
            MEMORY: 2.9 GB
             PANIC: "sysrq: SysRq : Trigger a crash"
               PID: 10635
           COMMAND: "bash"
              TASK: ffff8d6c84271800  [THREAD_INFO: ffff8d6c84271800]
               CPU: 1
             STATE: TASK_RUNNING (SYSRQ)
      
      crash>
  2. To exit the interactive prompt and stop crash, type exit or q.

    Example 17.2. Exiting the crash utility

    crash> exit
    ~]#
Note

The crash command can also be used as a powerful tool for debugging a live system. However use it with caution so as not to break your system.

17.3. Displaying various indicators in the crash utility

Use the crash utility to display various indicators, such as a kernel message buffer, a backtrace, a process status, virtual memory information and open files.

Displaying the message buffer

  • To display the kernel message buffer, type the log command at the interactive prompt:

    crash> log
    ... several lines omitted ...
    EIP: 0060:[<c068124f>] EFLAGS: 00010096 CPU: 2
    EIP is at sysrq_handle_crash+0xf/0x20
    EAX: 00000063 EBX: 00000063 ECX: c09e1c8c EDX: 00000000
    ESI: c0a09ca0 EDI: 00000286 EBP: 00000000 ESP: ef4dbf24
     DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
    Process bash (pid: 5591, ti=ef4da000 task=f196d560 task.ti=ef4da000)
    Stack:
     c068146b c0960891 c0968653 00000003 00000000 00000002 efade5c0 c06814d0
    <0> fffffffb c068150f b7776000 f2600c40 c0569ec4 ef4dbf9c 00000002 b7776000
    <0> efade5c0 00000002 b7776000 c0569e60 c051de50 ef4dbf9c f196d560 ef4dbfb4
    Call Trace:
     [<c068146b>] ? __handle_sysrq+0xfb/0x160
     [<c06814d0>] ? write_sysrq_trigger+0x0/0x50
     [<c068150f>] ? write_sysrq_trigger+0x3f/0x50
     [<c0569ec4>] ? proc_reg_write+0x64/0xa0
     [<c0569e60>] ? proc_reg_write+0x0/0xa0
     [<c051de50>] ? vfs_write+0xa0/0x190
     [<c051e8d1>] ? sys_write+0x41/0x70
     [<c0409adc>] ? syscall_call+0x7/0xb
    Code: a0 c0 01 0f b6 41 03 19 d2 f7 d2 83 e2 03 83 e0 cf c1 e2 04 09 d0 88 41 03 f3 c3 90 c7 05 c8 1b 9e c0 01 00 00 00 0f ae f8 89 f6 <c6> 05 00 00 00 00 01 c3 89 f6 8d bc 27 00 00 00 00 8d 50 d0 83
    EIP: [<c068124f>] sysrq_handle_crash+0xf/0x20 SS:ESP 0068:ef4dbf24
    CR2: 0000000000000000

    Type help log for more information about the command usage.

    Note

    The kernel message buffer includes the most essential information about the system crash and, as such, it is always dumped first in to the vmcore-dmesg.txt file. This is useful when an attempt to get the full vmcore file failed, for example because of lack of space on the target location. By default, vmcore-dmesg.txt is located in the /var/crash/ directory.

Displaying a backtrace

  • To display the kernel stack trace, use the bt command.

    crash> bt
    PID: 5591   TASK: f196d560  CPU: 2   COMMAND: "bash"
     #0 [ef4dbdcc] crash_kexec at c0494922
     #1 [ef4dbe20] oops_end at c080e402
     #2 [ef4dbe34] no_context at c043089d
     #3 [ef4dbe58] bad_area at c0430b26
     #4 [ef4dbe6c] do_page_fault at c080fb9b
     #5 [ef4dbee4] error_code (via page_fault) at c080d809
        EAX: 00000063  EBX: 00000063  ECX: c09e1c8c  EDX: 00000000  EBP: 00000000
        DS:  007b      ESI: c0a09ca0  ES:  007b      EDI: 00000286  GS:  00e0
        CS:  0060      EIP: c068124f  ERR: ffffffff  EFLAGS: 00010096
     #6 [ef4dbf18] sysrq_handle_crash at c068124f
     #7 [ef4dbf24] __handle_sysrq at c0681469
     #8 [ef4dbf48] write_sysrq_trigger at c068150a
     #9 [ef4dbf54] proc_reg_write at c0569ec2
    #10 [ef4dbf74] vfs_write at c051de4e
    #11 [ef4dbf94] sys_write at c051e8cc
    #12 [ef4dbfb0] system_call at c0409ad5
        EAX: ffffffda  EBX: 00000001  ECX: b7776000  EDX: 00000002
        DS:  007b      ESI: 00000002  ES:  007b      EDI: b7776000
        SS:  007b      ESP: bfcb2088  EBP: bfcb20b4  GS:  0033
        CS:  0073      EIP: 00edc416  ERR: 00000004  EFLAGS: 00000246

    Type bt <pid> to display the backtrace of a specific process or type help bt for more information about bt usage.

Displaying a process status

  • To display the status of processes in the system, use the ps command.

    crash> ps
       PID    PPID  CPU   TASK    ST  %MEM     VSZ    RSS  COMM
    >     0      0   0  c09dc560  RU   0.0       0      0  [swapper]
    >     0      0   1  f7072030  RU   0.0       0      0  [swapper]
          0      0   2  f70a3a90  RU   0.0       0      0  [swapper]
    >     0      0   3  f70ac560  RU   0.0       0      0  [swapper]
          1      0   1  f705ba90  IN   0.0    2828   1424  init
    ... several lines omitted ...
       5566      1   1  f2592560  IN   0.0   12876    784  auditd
       5567      1   2  ef427560  IN   0.0   12876    784  auditd
       5587   5132   0  f196d030  IN   0.0   11064   3184  sshd
    >  5591   5587   2  f196d560  RU   0.0    5084   1648  bash

    Use ps <pid> to display the status of a single specific process. Use help ps for more information about ps usage.

Displaying virtual memory information

  • To display basic virtual memory information, type the vm command at the interactive prompt.

    crash> vm
    PID: 5591   TASK: f196d560  CPU: 2   COMMAND: "bash"
       MM       PGD      RSS    TOTAL_VM
    f19b5900  ef9c6000  1648k    5084k
      VMA       START      END    FLAGS  FILE
    f1bb0310    242000    260000 8000875  /lib/ld-2.12.so
    f26af0b8    260000    261000 8100871  /lib/ld-2.12.so
    efbc275c    261000    262000 8100873  /lib/ld-2.12.so
    efbc2a18    268000    3ed000 8000075  /lib/libc-2.12.so
    efbc23d8    3ed000    3ee000 8000070  /lib/libc-2.12.so
    efbc2888    3ee000    3f0000 8100071  /lib/libc-2.12.so
    efbc2cd4    3f0000    3f1000 8100073  /lib/libc-2.12.so
    efbc243c    3f1000    3f4000 100073
    efbc28ec    3f6000    3f9000 8000075  /lib/libdl-2.12.so
    efbc2568    3f9000    3fa000 8100071  /lib/libdl-2.12.so
    efbc2f2c    3fa000    3fb000 8100073  /lib/libdl-2.12.so
    f26af888    7e6000    7fc000 8000075  /lib/libtinfo.so.5.7
    f26aff2c    7fc000    7ff000 8100073  /lib/libtinfo.so.5.7
    efbc211c    d83000    d8f000 8000075  /lib/libnss_files-2.12.so
    efbc2504    d8f000    d90000 8100071  /lib/libnss_files-2.12.so
    efbc2950    d90000    d91000 8100073  /lib/libnss_files-2.12.so
    f26afe00    edc000    edd000 4040075
    f1bb0a18   8047000   8118000 8001875  /bin/bash
    f1bb01e4   8118000   811d000 8101873  /bin/bash
    f1bb0c70   811d000   8122000 100073
    f26afae0   9fd9000   9ffa000 100073
    ... several lines omitted ...

    Use vm <pid> to display information about a single specific process, or use help vm for more information about vm usage.

Displaying open files

  • To display information about open files, use the files command.

    crash> files
    PID: 5591   TASK: f196d560  CPU: 2   COMMAND: "bash"
    ROOT: /    CWD: /root
     FD    FILE     DENTRY    INODE    TYPE  PATH
      0  f734f640  eedc2c6c  eecd6048  CHR   /pts/0
      1  efade5c0  eee14090  f00431d4  REG   /proc/sysrq-trigger
      2  f734f640  eedc2c6c  eecd6048  CHR   /pts/0
     10  f734f640  eedc2c6c  eecd6048  CHR   /pts/0
    255  f734f640  eedc2c6c  eecd6048  CHR   /pts/0

    Use files <pid> to display files opened by only one selected process, or use help files for more information about files usage.

17.4. Using Kernel Oops Analyzer

The Kernel Oops Analyzer tool analyzes the crash dump by comparing the oops messages with known issues in the knowledge base.

Prerequisites

  • Secure an oops message to feed the Kernel Oops Analyzer.

Procedure

  1. Access the Kernel Oops Analyzer tool.
  2. To diagnose a kernel crash issue, upload a kernel oops log generated in vmcore.

    • Alternatively you can also diagnose a kernel crash issue by providing a text message or a vmcore-dmesg.txt as an input.

      Kernel oops analyzer
  3. Click DETECT to compare the oops message based on information from the makedumpfile against known solutions.

17.5. The Kdump Helper tool

The Kdump Helper tool helps to set up the kdump using the provided information. Kdump Helper generates a configuration script based on your preferences. Initiating and running the script on your server sets up the kdump service.

Additional resources

Chapter 18. Using early kdump to capture boot time crashes

Early kdump is a feature of the kdump mechanism to capture the vmcore file if a system or kernel crash occurs during the early phases of the boot process before the system services start. Early kdump loads the crash kernel and the crash kernel’s initramfs in the memory much earlier.

18.1. What is early kdump

A kernel crash can sometimes occur during the early boot phase before the kdump service starts and is able to capture and save the contents of the crashed kernel memory. Therefore, crucial information related to the crash, which is important for troubleshooting, is lost. To address this problem, you can use the early kdump feature, which is part of the kdump service.

18.2. Enabling early kdump

The early kdump feature sets up the crash kernel and the initial RAM disk image (initramfs) to load early enough to capture the vmcore information for an early crash. This helps to eliminate the risk of losing information about the early boot kernel crashes.

Prerequisites

  • An active RHEL subscription.
  • A repository containing the kexec-tools package for your system CPU architecture.
  • Fulfilled kdump configuration and targets requirements. For more information see, Supported kdump configurations and targets.

Procedure

  1. Verify that the kdump service is enabled and active:

    # systemctl is-enabled kdump.service && systemctl is-active kdump.service
    enabled
    active

    If kdump is not enabled and running, set all required configurations and verify that kdump service is enabled.

  2. Rebuild the initramfs image of the booting kernel with the early kdump functionality:

    # dracut -f --add earlykdump
  3. Add the rd.earlykdump kernel command line parameter:

    # grubby --update-kernel=/boot/vmlinuz-$(uname -r) --args="rd.earlykdump"
  4. Reboot the system to reflect the changes

    # reboot

Verification step

  • Verify that rd.earlykdump was successfully added and early kdump feature was enabled:

    # cat /proc/cmdline
    BOOT_IMAGE=(hd0,msdos1)/vmlinuz-5.14.0-1.el9.x86_64 root=/dev/mapper/rhel-root ro crashkernel=auto resume=/dev/mapper/rhel-swap rd.lvm.lv=rhel/root rd.lvm.lv=rhel/swap rhgb quiet rd.earlykdump
    
    # journalctl -x | grep early-kdump
    Sep 13 15:46:11 redhat dracut-cmdline[304]: early-kdump is enabled.
    Sep 13 15:46:12 redhat dracut-cmdline[304]: kexec: loaded early-kdump kernel

Additional resources

Chapter 19. Signing a kernel and modules for Secure Boot

You can enhance the security of your system by using a signed kernel and signed kernel modules. On UEFI-based build systems where Secure Boot is enabled, you can self-sign a privately built kernel or kernel modules. Furthermore, you can import your public key into a target system where you want to deploy your kernel or kernel modules.

If Secure Boot is enabled, all of the following components have to be signed with a private key and authenticated with the corresponding public key:

  • UEFI operating system boot loader
  • The Red Hat Enterprise Linux kernel
  • All kernel modules

If any of these components are not signed and authenticated, the system cannot finish the booting process.

Red Hat Enterprise Linux 9 includes:

  • Signed boot loaders
  • Signed kernels
  • Signed kernel modules

In addition, the signed first-stage boot loader and the signed kernel include embedded Red Hat public keys. These signed executable binaries and embedded keys enable Red Hat Enterprise Linux 9 to install, boot, and run with the Microsoft UEFI Secure Boot Certification Authority keys that are provided by the UEFI firmware on systems that support UEFI Secure Boot.

Note
  • Not all UEFI-based systems include support for Secure Boot.
  • The build system, where you build and sign your kernel module, does not need to have UEFI Secure Boot enabled and does not even need to be a UEFI-based system.

19.1. Prerequisites

  • To be able to sign externally built kernel modules, install the utilities from the following packages:

    # dnf install pesign openssl kernel-devel mokutil keyutils

    Table 19.1. Required utilities

    UtilityProvided by packageUsed onPurpose

    efikeygen

    pesign

    Build system

    Generates public and private X.509 key pair

    openssl

    openssl

    Build system

    Exports the unencrypted private key

    sign-file

    kernel-devel

    Build system

    Executable file used to sign a kernel module with the private key

    mokutil

    mokutil

    Target system

    Optional utility used to manually enroll the public key

    keyctl

    keyutils

    Target system

    Optional utility used to display public keys in the system keyring

19.2. What is UEFI Secure Boot

With the Unified Extensible Firmware Interface (UEFI) Secure Boot technology, you can prevent the execution of the kernel-space code that has not been signed by a trusted key. The system boot loader is signed with a cryptographic key. The database of public keys, which is contained in the firmware, authorizes the signing key. You can subsequently verify a signature in the next-stage boot loader and the kernel.

UEFI Secure Boot establishes a chain of trust from the firmware to the signed drivers and kernel modules as follows:

  • An UEFI private key signs, and a public key authenticates the shim first-stage boot loader. A certificate authority (CA) in turn signs the public key. The CA is stored in the firmware database.
  • The shim file contains the Red Hat public key Red Hat Secure Boot (CA key 1) to authenticate the GRUB boot loader and the kernel.
  • The kernel in turn contains public keys to authenticate drivers and modules.

Secure Boot is the boot path validation component of the UEFI specification. The specification defines:

  • Programming interface for cryptographically protected UEFI variables in non-volatile storage.
  • Storing the trusted X.509 root certificates in UEFI variables.
  • Validation of UEFI applications such as boot loaders and drivers.
  • Procedures to revoke known-bad certificates and application hashes.

UEFI Secure Boot helps in the detection of unauthorized changes but does not:

  • Prevent installation or removal of second-stage boot loaders.
  • Require explicit user confirmation of such changes.
  • Stop boot path manipulations. Signatures are verified during booting, not when the boot loader is installed or updated.

If the boot loader or the kernel are not signed by a system trusted key, Secure Boot prevents them from starting.

19.3. UEFI Secure Boot support

You can install and run Red Hat Enterprise Linux 9 on systems with enabled UEFI Secure Boot if the kernel and all the loaded drivers are signed with a trusted key. Red Hat provides kernels and drivers that are signed and authenticated by the relevant Red Hat keys.

If you want to load externally built kernels or drivers, you must sign them as well.

Restrictions imposed by UEFI Secure Boot

  • The system only runs the kernel-mode code after its signature has been properly authenticated.
  • GRUB module loading is disabled because there is no infrastructure for signing and verification of GRUB modules. Allowing them to be loaded constitutes execution of untrusted code inside the security perimeter that Secure Boot defines.
  • Red Hat provides a signed GRUB binary that contains all the supported modules on Red Hat Enterprise Linux 9.

19.4. Requirements for authenticating kernel modules with X.509 keys

In Red Hat Enterprise Linux 9, when a kernel module is loaded, the kernel checks the signature of the module against the public X.509 keys from the kernel system keyring (.builtin_trusted_keys) and the kernel platform keyring (.platform). The .platform keyring contains keys from third-party platform providers and custom public keys. The keys from the kernel system .blacklist keyring are excluded from verification.

You need to meet certain conditions to load kernel modules on systems with enabled UEFI Secure Boot functionality:

  • If UEFI Secure Boot is enabled or if the module.sig_enforce kernel parameter has been specified:

    • You can only load those signed kernel modules whose signatures were authenticated against keys from the system keyring (.builtin_trusted_keys) and the platform keyring (.platform).
    • The public key must not be on the system revoked keys keyring (.blacklist).
  • If UEFI Secure Boot is disabled and the module.sig_enforce kernel parameter has not been specified:

    • You can load unsigned kernel modules and signed kernel modules without a public key.
  • If the system is not UEFI-based or if UEFI Secure Boot is disabled:

    • Only the keys embedded in the kernel are loaded onto .builtin_trusted_keys and .platform.
    • You have no ability to augment that set of keys without rebuilding the kernel.

Table 19.2. Kernel module authentication requirements for loading

Module signedPublic key found and signature validUEFI Secure Boot statesig_enforceModule loadKernel tainted

Unsigned

-

Not enabled

Not enabled

Succeeds

Yes

Not enabled

Enabled

Fails

-

Enabled

-

Fails

-

Signed

No

Not enabled

Not enabled

Succeeds

Yes

Not enabled

Enabled

Fails

-

Enabled

-

Fails

-

Signed

Yes

Not enabled

Not enabled

Succeeds

No

Not enabled

Enabled

Succeeds

No

Enabled

-

Succeeds

No

19.5. Sources for public keys

During boot, the kernel loads X.509 keys from a set of persistent key stores into the following keyrings:

  • The system keyring (.builtin_trusted_keys)
  • The .platform keyring
  • The system .blacklist keyring

Table 19.3. Sources for system keyrings

Source of X.509 keysUser can add keysUEFI Secure Boot stateKeys loaded during boot

Embedded in kernel

No

-

.builtin_trusted_keys

UEFI db

Limited

Not enabled

No

Enabled

.platform

Embedded in the shim boot loader

No

Not enabled

No

Enabled

.platform

Machine Owner Key (MOK) list

Yes

Not enabled

No

Enabled

.platform

.builtin_trusted_keys
  • A keyring that is built on boot
  • Contains trusted public keys
  • root privileges are needed to view the keys
.platform
  • A keyring that is built on boot
  • Contains keys from third-party platform providers and custom public keys
  • root privileges are needed to view the keys
.blacklist
  • A keyring with X.509 keys which have been revoked
  • A module signed by a key from .blacklist will fail authentication even if your public key is in .builtin_trusted_keys
UEFI Secure Boot db
  • A signature database
  • Stores keys (hashes) of UEFI applications, UEFI drivers, and boot loaders
  • The keys can be loaded on the machine
UEFI Secure Boot dbx
  • A revoked signature database
  • Prevents keys from being loaded
  • The revoked keys from this database are added to the .blacklist keyring

19.6. Generating a public and private key pair

To use a custom kernel or custom kernel modules on a Secure Boot-enabled system, you must generate a public and private X.509 key pair. You can use the generated private key to sign the kernel or the kernel modules. You can also validate the signed kernel or kernel modules by adding the corresponding public key to the Machine Owner Key (MOK) for Secure Boot.

Warning

Apply strong security measures and access policies to guard the contents of your private key. In the wrong hands, the key could be used to compromise any system which is authenticated by the corresponding public key.

Procedure

  • Create an X.509 public and private key pair:

    • If you only want to sign custom kernel modules:

      # efikeygen --dbdir /etc/pki/pesign \
                  --self-sign \
                  --module \
                  --common-name 'CN=Organization signing key' \
                  --nickname 'Custom Secure Boot key'
    • If you want to sign custom kernel:

      # efikeygen --dbdir /etc/pki/pesign \
                  --self-sign \
                  --kernel \
                  --common-name 'CN=Organization signing key' \
                  --nickname 'Custom Secure Boot key'
    • When the RHEL system is running FIPS mode:

      # efikeygen --dbdir /etc/pki/pesign \
                  --self-sign \
                  --kernel \
                  --common-name 'CN=Organization signing key' \
                  --nickname 'Custom Secure Boot key'
                  --token 'NSS FIPS 140-2 Certificate DB'
      Note

      In FIPS mode, you must use the --token option so that efikeygen finds the default "NSS Certificate DB" token in the PKI database.

      The public and private keys are now stored in the /etc/pki/pesign/ directory.

Important

It is a good security practice to sign the kernel and the kernel modules within the validity period of its signing key. However, the sign-file utility does not warn you and the key will be usable in Red Hat Enterprise Linux 9 regardless of the validity dates.

19.7. Example output of system keyrings

You can display information about the keys on the system keyrings using the keyctl utility from the keyutils package.

Prerequisites

  • You have root permissions.
  • You have installed the keyctl utility from the keyutils package.

Example 19.1. Keyrings output

The following is a shortened example output of .builtin_trusted_keys, .platform, and .blacklist keyrings from a Red Hat Enterprise Linux 9 system where UEFI Secure Boot is enabled.

# keyctl list %:.builtin_trusted_keys
6 keys in keyring:
...asymmetric: Red Hat Enterprise Linux Driver Update Program (key 3): bf57f3e87...
...asymmetric: Red Hat Secure Boot (CA key 1): 4016841644ce3a810408050766e8f8a29...
...asymmetric: Microsoft Corporation UEFI CA 2011: 13adbf4309bd82709c8cd54f316ed...
...asymmetric: Microsoft Windows Production PCA 2011: a92902398e16c49778cd90f99e...
...asymmetric: Red Hat Enterprise Linux kernel signing key: 4249689eefc77e95880b...
...asymmetric: Red Hat Enterprise Linux kpatch signing key: 4d38fd864ebe18c5f0b7...

# keyctl list %:.platform
4 keys in keyring:
...asymmetric: VMware, Inc.: 4ad8da0472073...
...asymmetric: Red Hat Secure Boot CA 5: cc6fafe72...
...asymmetric: Microsoft Windows Production PCA 2011: a929f298e1...
...asymmetric: Microsoft Corporation UEFI CA 2011: 13adbf4e0bd82...

# keyctl list %:.blacklist
4 keys in keyring:
...blacklist: bin:f5ff83a...
...blacklist: bin:0dfdbec...
...blacklist: bin:38f1d22...
...blacklist: bin:51f831f...

The .builtin_trusted_keys keyring in the example shows the addition of two keys from the UEFI Secure Boot db keys as well as the Red Hat Secure Boot (CA key 1), which is embedded in the shim boot loader.

Example 19.2. Kernel console output

The following example shows the kernel console output. The messages identify the keys with an UEFI Secure Boot related source. These include UEFI Secure Boot db, embedded shim, and MOK list.

# dmesg | egrep 'integrity.*cert'
[1.512966] integrity: Loading X.509 certificate: UEFI:db
[1.513027] integrity: Loaded X.509 cert 'Microsoft Windows Production PCA 2011: a929023...
[1.513028] integrity: Loading X.509 certificate: UEFI:db
[1.513057] integrity: Loaded X.509 cert 'Microsoft Corporation UEFI CA 2011: 13adbf4309...
[1.513298] integrity: Loading X.509 certificate: UEFI:MokListRT (MOKvar table)
[1.513549] integrity: Loaded X.509 cert 'Red Hat Secure Boot CA 5: cc6fa5e72868ba494e93...

Additional resources

  • keyctl(1), dmesg(1) manual pages

19.8. Enrolling public key on target system by adding the public key to the MOK list

You must enroll your public key on all systems where you want to authenticate and load your kernel or kernel modules. You can import the public key on a target system in different ways so that the platform keyring (.platform) is able to use the public key to authenticate the kernel or kernel modules.

When RHEL 9 boots on a UEFI-based system with Secure Boot enabled, the kernel loads onto the platform keyring (.platform) all public keys that are in the Secure Boot db key database. At the same time, the kernel excludes the keys in the dbx database of revoked keys.

You can use the Machine Owner Key (MOK) facility feature to expand the UEFI Secure Boot key database. When RHEL 9 boots on an UEFI-enabled system with Secure Boot enabled, the keys on the MOK list are also added to the platform keyring (.platform) in addition to the keys from the key database. The MOK list keys are also stored persistently and securely in the same fashion as the Secure Boot database keys, but these are two separate facilities. The MOK facility is supported by shim, MokManager, GRUB, and the mokutil utility.

Note

To facilitate authentication of your kernel module on your systems, consider requesting your system vendor to incorporate your public key into the UEFI Secure Boot key database in their factory firmware image.

Prerequisites

Procedure

  1. Export your public key to the sb_cert.cer file:

    # certutil -d /etc/pki/pesign \
               -n 'Custom Secure Boot key' \
               -Lr \
               > sb_cert.cer
  2. Import your public key into the MOK list:

    # mokutil --import sb_cert.cer
  3. Enter a new password for this MOK enrollment request.
  4. Reboot the machine.

    The shim boot loader notices the pending MOK key enrollment request and it launches MokManager.efi to enable you to complete the enrollment from the UEFI console.

  5. Choose Enroll MOK, enter the password you previously associated with this request when prompted, and confirm the enrollment.

    Your public key is added to the MOK list, which is persistent.

    Once a key is on the MOK list, it will be automatically propagated to the .platform keyring on this and subsequent boots when UEFI Secure Boot is enabled.

19.9. Signing a kernel with the private key

You can obtain enhanced security benefits on your system by loading a signed kernel if the UEFI Secure Boot mechanism is enabled.

Prerequisites

Procedure

  • On the x64 architecture:

    1. Create a signed image:

      # pesign --certificate 'Custom Secure Boot key' \
               --in vmlinuz-version \
               --sign \
               --out vmlinuz-version.signed

      Replace version with the version suffix of your vmlinuz file, and Custom Secure Boot key with the name that you chose earlier.

    2. Optional: Check the signatures:

      # pesign --show-signature \
               --in vmlinuz-version.signed
    3. Overwrite the unsigned image with the signed image:

      # mv vmlinuz-version.signed vmlinuz-version
  • On the 64-bit ARM architecture:

    1. Decompress the vmlinuz file:

      # zcat vmlinuz-version > vmlinux-version
    2. Create a signed image:

      # pesign --certificate 'Custom Secure Boot key' \
               --in vmlinux-version \
               --sign \
               --out vmlinux-version.signed
    3. Optional: Check the signatures:

      # pesign --show-signature \
               --in vmlinux-version.signed
    4. Compress the vmlinux file:

      # gzip --to-stdout vmlinux-version.signed > vmlinuz-version
    5. Remove the uncompressed vmlinux file:

      # rm vmlinux-version*

19.10. Signing a GRUB build with the private key

On a system where the UEFI Secure Boot mechanism is enabled, you can sign a GRUB build with a custom existing private key. You must do this if you are using a custom GRUB build, or if you have removed the Microsoft trust anchor from your system.

Prerequisites

Procedure

  • On the x64 architecture:

    1. Create a signed GRUB EFI binary:

      # pesign --in /boot/efi/EFI/redhat/grubx64.efi \
               --out /boot/efi/EFI/redhat/grubx64.efi.signed \
               --certificate 'Custom Secure Boot key' \
               --sign

      Replace Custom Secure Boot key with the name that you chose earlier.

    2. Optional: Check the signatures:

      # pesign --in /boot/efi/EFI/redhat/grubx64.efi.signed \
               --show-signature
    3. Overwrite the unsigned binary with the signed binary:

      # mv /boot/efi/EFI/redhat/grubx64.efi.signed \
           /boot/efi/EFI/redhat/grubx64.efi
  • On the 64-bit ARM architecture:

    1. Create a signed GRUB EFI binary:

      # pesign --in /boot/efi/EFI/redhat/grubaa64.efi \
               --out /boot/efi/EFI/redhat/grubaa64.efi.signed \
               --certificate 'Custom Secure Boot key' \
               --sign

      Replace Custom Secure Boot key with the name that you chose earlier.

    2. Optional: Check the signatures:

      # pesign --in /boot/efi/EFI/redhat/grubaa64.efi.signed \
               --show-signature
    3. Overwrite the unsigned binary with the signed binary:

      # mv /boot/efi/EFI/redhat/grubaa64.efi.signed \
           /boot/efi/EFI/redhat/grubaa64.efi

19.11. Signing kernel modules with the private key

You can enhance the security of your system by loading signed kernel modules if the UEFI Secure Boot mechanism is enabled.

Your signed kernel module is also loadable on systems where UEFI Secure Boot is disabled or on a non-UEFI system. As a result, you do not need to provide both a signed and unsigned version of your kernel module.

Prerequisites

Procedure

  1. Export your public key to the sb_cert.cer file:

    # certutil -d /etc/pki/pesign \
               -n 'Custom Secure Boot key' \
               -Lr \
               > sb_cert.cer
  2. Extract the key from the NSS database as a PKCS #12 file:

    # pk12util -o sb_cert.p12 \
               -n 'Custom Secure Boot key' \
               -d /etc/pki/pesign
  3. When the previous command prompts you, enter a new password that encrypts the private key.
  4. Export the unencrypted private key:

    # openssl pkcs12 \
             -in sb_cert.p12 \
             -out sb_cert.priv \
             -nocerts \
             -noenc
    Important

    Handle the unencrypted private key with care.

  5. Sign your kernel module. The following command appends the signature directly to the ELF image in your kernel module file:

    # /usr/src/kernels/$(uname -r)/scripts/sign-file \
              sha256 \
              sb_cert.priv \
              sb_cert.cer \
              my_module.ko

Your kernel module is now ready for loading.

Important

In Red Hat Enterprise Linux 9, the validity dates of the key pair matter. The key does not expire, but the kernel module must be signed within the validity period of its signing key. The sign-file utility will not warn you of this. For example, a key that is only valid in 2021 can be used to authenticate a kernel module signed in 2021 with that key. However, users cannot use that key to sign a kernel module in 2022.

Verification

  1. Display information about the kernel module’s signature:

    # modinfo my_module.ko | grep signer
      signer:         Your Name Key

    Check that the signature lists your name as entered during generation.

    Note

    The appended signature is not contained in an ELF image section and is not a formal part of the ELF image. Therefore, utilities such as readelf cannot display the signature on your kernel module.

  2. Load the module:

    # insmod my_module.ko
  3. Remove (unload) the module:

    # modprobe -r my_module.ko

19.12. Loading signed kernel modules

Once your public key is enrolled in the system keyring (.builtin_trusted_keys) and the MOK list, and after you have signed the respective kernel module with your private key, you can load your signed kernel module with the modprobe command.

Prerequisites

Procedure

  1. Verify that your public keys are on the system keyring:

    # keyctl list %:.platform
  2. Copy the kernel module into the extra/ directory of the kernel that you want:

    # cp my_module.ko /lib/modules/$(uname -r)/extra/
  3. Update the modular dependency list:

    # depmod -a
  4. Load the kernel module:

    # modprobe -v my_module
  5. Optionally, to load the module on boot, add it to the /etc/modules-loaded.d/my_module.conf file:

    # echo "my_module" > /etc/modules-load.d/my_module.conf

Verification

  • Verify that the module was successfully loaded:

    # lsmod | grep my_module

Additional resources

Chapter 20. Updating the Secure Boot Revocation List

You can update the UEFI Secure Boot Revocation List on your system so that Secure Boot identifies software with known security issues and prevents it from compromising your boot process.

20.1. Prerequisites

  • Secure Boot is enabled on your system.

20.2. What is UEFI Secure Boot

With the Unified Extensible Firmware Interface (UEFI) Secure Boot technology, you can prevent the execution of the kernel-space code that has not been signed by a trusted key. The system boot loader is signed with a cryptographic key. The database of public keys, which is contained in the firmware, authorizes the signing key. You can subsequently verify a signature in the next-stage boot loader and the kernel.

UEFI Secure Boot establishes a chain of trust from the firmware to the signed drivers and kernel modules as follows:

  • An UEFI private key signs, and a public key authenticates the shim first-stage boot loader. A certificate authority (CA) in turn signs the public key. The CA is stored in the firmware database.
  • The shim file contains the Red Hat public key Red Hat Secure Boot (CA key 1) to authenticate the GRUB boot loader and the kernel.
  • The kernel in turn contains public keys to authenticate drivers and modules.

Secure Boot is the boot path validation component of the UEFI specification. The specification defines:

  • Programming interface for cryptographically protected UEFI variables in non-volatile storage.
  • Storing the trusted X.509 root certificates in UEFI variables.
  • Validation of UEFI applications such as boot loaders and drivers.
  • Procedures to revoke known-bad certificates and application hashes.

UEFI Secure Boot helps in the detection of unauthorized changes but does not:

  • Prevent installation or removal of second-stage boot loaders.
  • Require explicit user confirmation of such changes.
  • Stop boot path manipulations. Signatures are verified during booting, not when the boot loader is installed or updated.

If the boot loader or the kernel are not signed by a system trusted key, Secure Boot prevents them from starting.

20.3. The Secure Boot Revocation List

The UEFI Secure Boot Revocation List, or the Secure Boot Forbidden Signature Database (dbx), is a list that identifies software that Secure Boot no longer allows to run.

When a security issue or a stability problem is found in software that interfaces with Secure Boot, such as in the GRUB boot loader, the Revocation List stores its hash signature. Software with such a recognized signature cannot run during boot, and the system boot fails to prevent compromising the system.

For example, a certain version of GRUB might contain a security issue that allows an attacker to bypass the Secure Boot mechanism. When the issue is found, the Revocation List adds hash signatures of all GRUB versions that contain the issue. As a result, only secure GRUB versions can boot on the system.

The Revocation List requires regular updates to recognize newly found issues. When updating the Revocation List, make sure to use a safe update method that does not cause your currently installed system to no longer boot.

20.4. Applying an online Revocation List update

You can update the Secure Boot Revocation List on your system so that Secure Boot prevents known security issues. This procedure is safe and ensures that the update does not prevent your system from booting.

Prerequisites

  • Your system can access the internet for updates.

Procedure

  1. Determine the current version of the Revocation List:

    # fwupdmgr get-devices

    See the Current version field under UEFI dbx.

  2. Enable the LVFS Revocation List repository:

    # fwupdmgr enable-remote lvfs
  3. Refresh the repository metadata:

    # fwupdmgr refresh
  4. Apply the Revocation List update:

    • On the command line:

      # fwupdmgr update
    • In the graphical interface:

      1. Open the Software application
      2. Navigate to the Updates tab.
      3. Find the Secure Boot dbx Configuration Update entry.
      4. Click Update.
  5. At the end of the update, fwupdmgr or Software asks you to reboot the system. Confirm the reboot.

Verification

  • After the reboot, check the current version of the Revocation List again:

    # fwupdmgr get-devices

20.5. Applying an offline Revocation List update

On a system with no internet connection, you can update the Secure Boot Revocation List from RHEL so that Secure Boot prevents known security issues. This procedure is safe and ensures that the update does not prevent your system from booting.

Procedure

  1. Determine the current version of the Revocation List:

    # fwupdmgr get-devices

    See the Current version field under UEFI dbx.

  2. List the updates available from RHEL:

    # ls /usr/share/dbxtool/
  3. Select the most recent update file for your architecture. The file names use the following format:

    DBXUpdate-date-architecture.cab
  4. Install the selected update file:

    # fwupdmgr install /usr/share/dbxtool/DBXUpdate-date-architecture.cab
  5. At the end of the update, fwupdmgr asks you to reboot the system. Confirm the reboot.

Verification

  • After the reboot, check the current version of the Revocation List again:

    # fwupdmgr get-devices

Chapter 21. Enhancing security with the kernel integrity subsystem

You can improve the protection of your system by using components of the kernel integrity subsystem. Learn more about the relevant components and their configuration.

Note

You can use the features with cryptographic signatures only for Red Hat products because the kernel keyring system includes only the certificates for Red Hat signature keys. Using other hash features results in incomplete tamper-proofing.

21.1. The kernel integrity subsystem

The integrity subsystem is the part of the kernel that maintains overall integrity of system data. This subsystem helps to keep the state of a system the same from the time it was built. By using this subsystem, you can prevent undesired modification of specific system files.

The kernel integrity subsystem consists of two major components:

Integrity Measurement Architecture (IMA)
  • IMA measures file content whenever it is executed or opened by cryptographically hashing or signing with cryptographic keys. The keys are stored in the kernel keyring subsystem.
  • IMA places the measured values within the kernel’s memory space. This prevents users of the system from modifying the measured values.
  • IMA allows local and remote parties to verify the measured values.
  • IMA provides local validation of the current content of files against the values previously stored in the measurement list within the kernel memory. This extension forbids performing any operation on a specific file in case the current and the previous measures do not match.
Extended Verification Module (EVM)
  • EVM protects extended attributes of files (also known as xattr) that are related to system security, such as IMA measurements and SELinux attributes. EVM cryptographically hashes their corresponding values or signs them with cryptographic keys. The keys are stored in the kernel keyring subsystem.

The kernel integrity subsystem can use the Trusted Platform Module (TPM) to further harden system security.

A TPM is a hardware, firmware, or virtual component with integrated cryptographic keys, which is built according to the TPM specification by the Trusted Computing Group (TCG) for important cryptographic functions. TPMs are usually built as dedicated hardware attached to the platform’s motherboard. By providing cryptographic functions from a protected and tamper-proof area of the hardware chip, TPMs are protected from software-based attacks. TPMs provide the following features:

  • Random-number generator
  • Generator and secure storage for cryptographic keys
  • Hashing generator
  • Remote attestation

21.2. Trusted and encrypted keys

Trusted keys and encrypted keys are an important part of enhancing system security.

Trusted and encrypted keys are variable-length symmetric keys generated by the kernel that use the kernel keyring service. The integrity of the keys can be verified, which means that they can be used, for example, by the extended verification module (EVM) to verify and confirm the integrity of a running system. User-level programs can only access the keys in the form of encrypted blobs.

Trusted keys

Trusted keys need the Trusted Platform Module (TPM) chip, which is used to both create and encrypt (seal) the keys. Each TPM has a master wrapping key, called the storage root key, which is stored within the TPM itself.

Note

RHEL 9 supports only TPM 2.0. If you must use TPM 1.2, use RHEL 8. For more information, see the Is Trusted Platform Module (TPM) supported by Red Hat? solution.

You can verify that a TPM 2.0 chip has been enabled by entering the following command:

$ cat /sys/class/tpm/tpm0/tpm_version_major
2

You can also enable a TPM 2.0 chip and manage the TPM 2.0 device through settings in the machine firmware.

In addition to that, you can seal the trusted keys with a specific set of the TPM’s platform configuration register (PCR) values. PCR contains a set of integrity-management values that reflect the firmware, boot loader, and operating system. This means that PCR-sealed keys can only be decrypted by the TPM on the same system on which they were encrypted. However, when a PCR-sealed trusted key is loaded (added to a keyring), and thus its associated PCR values are verified, it can be updated with new (or future) PCR values, so that a new kernel, for example, can be booted. You can save a single key also as multiple blobs, each with a different PCR value.

Encrypted keys
Encrypted keys do not require a TPM, because they use the kernel Advanced Encryption Standard (AES), which makes them faster than trusted keys. Encrypted keys are created using kernel-generated random numbers and encrypted by a master key when they are exported into user-space blobs.

The master key is either a trusted key or a user key. If the master key is not trusted, the encrypted key is only as secure as the user key used to encrypt it.

21.3. Working with trusted keys

You can improve system security by using the keyctl utility to create, export, load and update trusted keys.

Prerequisites

Procedure

  1. Create a 2048-bit RSA key with an SHA-256 primary storage key with a persistent handle of, for example, 81000001, by using one of the following utilities:

    1. By using the tss2 package:

      # TPM_DEVICE=/dev/tpm0 tsscreateprimary -hi o -st
      Handle 80000000
      # TPM_DEVICE=/dev/tpm0 tssevictcontrol -hi o -ho 80000000 -hp 81000001
    2. By using the tpm2-tools package:

      # tpm2_createprimary --key-algorithm=rsa2048 --key-context=key.ctxt
      name-alg:
        value: sha256
        raw: 0xb
      …
      sym-keybits: 128
      rsa: xxxxxx…
      
      # tpm2_evictcontrol -c key.ctxt 0x81000001
      persistentHandle: 0x81000001
      action: persisted
  2. Create a trusted key by using a TPM 2.0 with the syntax of keyctl add trusted <NAME> "new <KEY_LENGTH> keyhandle=<PERSISTENT-HANDLE> [options]" <KEYRING>. In this example, the persistent handle is 81000001.

    # keyctl add trusted kmk "new 32 keyhandle=0x81000001" @u
    642500861

    The command creates a trusted key called kmk with the length of 32 bytes (256 bits) and places it in the user keyring (@u). The keys may have a length of 32 to 128 bytes (256 to 1024 bits).

  3. List the current structure of the kernel keyrings:

    # keyctl show
    Session Keyring
           -3 --alswrv    500   500  keyring: ses 97833714 --alswrv 500 -1 \ keyring: uid.1000 642500861 --alswrv 500 500 \ trusted: kmk
  4. Export the key to a user-space blob by using the serial number of the trusted key:

    # keyctl pipe 642500861 > kmk.blob

    The command uses the pipe subcommand and the serial number of kmk.

  5. Load the trusted key from the user-space blob:

    # keyctl add trusted kmk "load `cat kmk.blob`" @u
    268728824
  6. Create secure encrypted keys that use the TPM-sealed trusted key (kmk). Follow this syntax: keyctl add encrypted <NAME> "new [FORMAT] <KEY_TYPE>:<PRIMARY_KEY_NAME> <KEY_LENGTH>" <KEYRING>:

    # keyctl add encrypted encr-key "new trusted:kmk 32" @u
    159771175

21.4. Working with encrypted keys

You can improve system security on systems where a Trusted Platform Module (TPM) is not available by managing encrypted keys.

Procedure

  1. Generate a user key by using a random sequence of numbers.

    # keyctl add user kmk-user "$(dd if=/dev/urandom bs=1 count=32 2>/dev/null)" @u
    427069434

    The command generates a user key called kmk-user which acts as a primary key and is used to seal the actual encrypted keys.

  2. Generate an encrypted key using the primary key from the previous step:

    # keyctl add encrypted encr-key "new user:kmk-user 32" @u
    1012412758
  3. Optionally, list all keys in the specified user keyring:

    # keyctl list @u
    2 keys in keyring:
    427069434: --alswrv  1000  1000 user: kmk-user
    1012412758: --alswrv  1000  1000 encrypted: encr-key
Important

Encrypted keys that are not sealed by a trusted primary key are only as secure as the user primary key (random-number key) that was used to encrypt them. Therefore, load the primary user key as securely as possible and preferably early during the boot process.

Additional resources

21.5. Enabling IMA and EVM

You can enable and configure Integrity measurement architecture (IMA) and extended verification module (EVM) to improve the security of the operating system.

Prerequisites

  • Secure Boot is temporarily disabled.

    Note

    When Secure Boot is enabled, the ima_appraise=fix kernel command-line parameter does not work.

  • The securityfs file system is mounted on the /sys/kernel/security/ directory and the /sys/kernel/security/integrity/ima/ directory exists. You can verify where securityfs is mounted by using the mount command:

    # mount
    ...
    securityfs on /sys/kernel/security type securityfs (rw,nosuid,nodev,noexec,relatime)
    ...
  • The systemd service manager is patched to support IMA and EVM on boot time. Verify by using the following command:

    # grep <options> pattern <files>

    For example:

    # dmesg | grep -i -e EVM -e IMA -w
    [ 0.943873] ima: No TPM chip found, activating TPM-bypass!
    [ 0.944566] ima: Allocated hash algorithm: sha256
    [ 0.944579] ima: No architecture policies found
    [ 0.944601] evm: Initialising EVM extended attributes:
    [ 0.944602] evm: security.selinux
    [ 0.944604] evm: security.SMACK64 (disabled)
    [ 0.944605] evm: security.SMACK64EXEC (disabled)
    [ 0.944607] evm: security.SMACK64TRANSMUTE (disabled)
    [ 0.944608] evm: security.SMACK64MMAP (disabled)
    [ 0.944609] evm: security.apparmor (disabled)
    [ 0.944611] evm: security.ima
    [ 0.944612] evm: security.capability
    [ 0.944613] evm: HMAC attrs: 0x1
    [ 1.314520] systemd[1]: systemd 252-18.el9 running in system mode (+PAM +AUDIT +SELINUX -APPARMOR +IMA +SMACK +SECCOMP +GCRYPT +GNUTLS +OPENSSL +ACL +BLKID +CURL +ELFUTILS -FIDO2 +IDN2 -IDN -IPTC +KMOD +LIBCRYPTSETUP +LIBFDISK +PCRE2 -PWQUALITY +P11KIT -QRENCODE +TPM2 +BZIP2 +LZ4 +XZ +ZLIB +ZSTD -BPF_FRAMEWORK +XKBCOMMON +UTMP +SYSVINIT default-hierarchy=unified)
    [ 1.717675] device-mapper: core: CONFIG_IMA_DISABLE_HTABLE is disabled. Duplicate IMA measurements will not be recorded in the IMA log.
    [ 4.799436] systemd[1]: systemd 252-18.el9 running in system mode (+PAM +AUDIT +SELINUX -APPARMOR +IMA +SMACK +SECCOMP +GCRYPT +GNUTLS +OPENSSL +ACL +BLKID +CURL +ELFUTILS -FIDO2 +IDN2 -IDN -IPTC +KMOD +LIBCRYPTSETUP +LIBFDISK +PCRE2 -PWQUALITY +P11KIT -QRENCODE +TPM2 +BZIP2 +LZ4 +XZ +ZLIB +ZSTD -BPF_FRAMEWORK +XKBCOMMON +UTMP +SYSVINIT default-hierarchy=unified)

Procedure

  1. Enable IMA and EVM in the fix mode for the current boot entry and allow users to gather and update the IMA measurements by adding the following kernel command-line parameters:

    # grubby --update-kernel=/boot/vmlinuz-$(uname -r) --args="ima_policy=appraise_tcb ima_appraise=fix evm=fix"

    The command enables IMA and EVM in the fix mode for the current boot entry and allows users to gather and update the IMA measurements.

    The ima_policy=appraise_tcb kernel command-line parameter ensures that the kernel uses the default Trusted Computing Base (TCB) measurement policy and the appraisal step. The appraisal step forbids access to files whose prior and current measures do not match.

  2. Reboot to make the changes come into effect.
  3. Optional: Verify that the parameters have been added to the kernel command line:

    # cat /proc/cmdline
    BOOT_IMAGE=(hd0,msdos1)/vmlinuz-5.14.0-1.el9.x86_64 root=/dev/mapper/rhel-root ro crashkernel=1G-4G:192M,4G-64G:256M,64G-:512M resume=/dev/mapper/rhel-swap rd.lvm.lv=rhel/root rd.lvm.lv=rhel/swap rhgb quiet ima_policy=appraise_tcb ima_appraise=fix evm=fix
  4. Create a kernel master key to protect the EVM key:

    # keyctl add user kmk "$(dd if=/dev/urandom bs=1 count=32 2> /dev/null)" @u
    748544121

    The kmk is kept entirely in the kernel space memory. The 32-byte long value of the kmk is generated from random bytes from the /dev/urandom file and placed in the user (@u) keyring. The key serial number is on the first line of the previous output.

  5. Create an encrypted EVM key based on the kmk:

    # keyctl add encrypted evm-key "new user:kmk 64" @u
    641780271

    The command uses the kmk to generate and encrypt a 64-byte long user key (named evm-key) and places it in the user (@u) keyring. The key serial number is on the first line of the previous output.

    Important

    It is necessary to name the user key as evm-key because that is the name the EVM subsystem is expecting and is working with.

  6. Create a directory for exported keys.

    # mkdir -p /etc/keys/
  7. Search for the kmk and export its unencrypted value into the new directory.

    # keyctl pipe $(keyctl search @u user kmk) > /etc/keys/kmk
  8. Search for the evm-key and export its encrypted value into the new directory.

    # keyctl pipe $(keyctl search @u encrypted evm-key) > /etc/keys/evm-key

    The evm-key has been encrypted by the kernel master key earlier.

  9. Optional: View the newly created keys.

    # keyctl show
    Session Keyring
    974575405   --alswrv     0        0      keyring: ses 299489774 --alswrv 0 65534 \ keyring: uid.0 748544121 --alswrv 0 0 \ user: kmk
    641780271   --alswrv     0        0           \_ encrypted: evm-key
    
    # ls -l /etc/keys/
    total 8
    -rw-r--r--. 1 root root 246 Jun 24 12:44 evm-key
    -rw-r--r--. 1 root root  32 Jun 24 12:43 kmk
  10. Optional: If the keys have been removed from the keyring, for example after system reboot, you can import the already exported kmk and evm-key instead of creating new ones.

    1. Import the kmk.

      # keyctl add user kmk "$(cat /etc/keys/kmk)" @u
      451342217
    2. Import the evm-key.

      # keyctl add encrypted evm-key "load $(cat /etc/keys/evm-key)" @u
      924537557
  11. Activate EVM.

    # echo 1 > /sys/kernel/security/evm
  12. Relabel the whole system.

    # find / -fstype xfs -type f -uid 0 -exec head -n 1 '{}' >/dev/null \;
    Warning

    Enabling IMA and EVM without relabeling the system might make the majority of the files on the system inaccessible.

Verification

  • Verify that EVM has been initialized.

    # dmesg | tail -1
    […​] evm: key initialized

21.6. Collecting file hashes with integrity measurement architecture

In the measurement phase, you can create file hashes and store them as extended attributes (xattrs) of those files. With the file hashes, you can generate either an RSA-based digital signature or a Hash-based Message Authentication Code (HMAC-SHA1) and thus prevent offline tampering attacks on the extended attributes.

Prerequisites

Procedure

  1. Create a test file:

    # echo <Test_text> > test_file

    IMA and EVM ensure that the test_file example file has assigned hash values that are stored as its extended attributes.

  2. Inspect the file’s extended attributes:

    # getfattr -m . -d test_file
    # file: test_file
    security.evm=0sAnDIy4VPA0HArpPO/EqiutnNyBql
    security.ima=0sAQOEDeuUnWzwwKYk+n66h/vby3eD

    The example output shows extended attributes with the IMA and EVM hash values and SELinux context. EVM adds a security.evm extended attribute related to the other attributes. At this point, you can use the evmctl utility on security.evm to generate either an RSA-based digital signature or a Hash-based Message Authentication Code (HMAC-SHA1).

Additional resources

21.7. Adding IMA signatures to package files

You need to add IMA signatures to RPM files to allow the kernel, Keylime, fapolicyd, and debuginfo packages to do their integrity checks. After installing the rpm-plugin-ima plugin, newly installed RPM files automatically have IMA signatures placed in the security.ima extended file attribute. However, you need to reinstall existing packages to obtain IMA signatures.

Procedure

  1. To install the rpm-plugin-ima plugin, run:

    # dnf install rpm-plugin-ima -y
  2. To reinstall all packages, run:

    # dnf reinstall “*” -y

Verification

  1. Confirm that the reinstalled package file has the valid IMA signature. For example, to check the IMA signature of the /usr/bin/bash file, run:

    # getfattr -m security.ima -d /usr/bin/bash
    security.ima=0sAwIE0zIESQBnMGUCMFhf0iBeM7NjjhCCHVt4/ORx1eCegjrWSHzFbJMCsAhR9bYU2hNGjiWUYT2IIqWaaAIxALFGUkqGP5vDLuxQXibO9g7HFcfyZzRBY4rbKPsXcAIZRtDHVS5dQBZqM3hyS5v1MA==
  2. Verify IMA signature of a file with a specified certificate. The IMA code signing key is accessible by /usr/share/doc/kernel-keys/$(uname -r)/ima.cer.

    # evmctl ima_verify -k /usr/share/doc/kernel-keys/$(uname -r)/ima.cer /usr/bin/bash
    key 1: d3320449 /usr/share/doc/kernel-keys/5.14.0-359.el9.x86-64/ima.cer
    /usr/bin/bash: verification is OK

21.8. Enabling kernel runtime integrity monitoring

You can enable kernel runtime integrity monitoring that IMA appraisal provides.

Prerequisites

  • The kernel installed on your system has version 5.14.0-359 or higher.
  • The dracut package has version 057-43.git20230816 or higher.
  • The keyutils package is installed.
  • The ima-evm-utils package is installed.
  • The files covered by the policy have valid signatures. For instructions, see Adding IMA signatures to package files.

Procedure

  1. To copy the Red Hat IMA code signing key to the /etc/ima/keys file, run:

    $ mkdir -p /etc/keys/ima
    $ cp /usr/share/doc/kernel-keys/$(uname -r)/ima.cer /etc/ima/keys
  2. To add the IMA code signing key to the .ima keyring, run:

    # keyctl padd asymmetric RedHat-IMA %:.ima < /etc/ima/keys/ima.cer
  3. Depending on your threat model, define an IMA policy in the /etc/sysconfig/ima-policy file. For example, the following IMA policy checks the integrity of both executables and involved memory mapping library files:

    # PROC_SUPER_MAGIC = 0x9fa0
    dont_appraise fsmagic=0x9fa0
    # SYSFS_MAGIC = 0x62656572
    dont_appraise fsmagic=0x62656572
    # DEBUGFS_MAGIC = 0x64626720
    dont_appraise fsmagic=0x64626720
    # TMPFS_MAGIC = 0x01021994
    dont_appraise fsmagic=0x1021994
    # RAMFS_MAGIC
    dont_appraise fsmagic=0x858458f6
    # DEVPTS_SUPER_MAGIC=0x1cd1
    dont_appraise fsmagic=0x1cd1
    # BINFMTFS_MAGIC=0x42494e4d
    dont_appraise fsmagic=0x42494e4d
    # SECURITYFS_MAGIC=0x73636673
    dont_appraise fsmagic=0x73636673
    # SELINUX_MAGIC=0xf97cff8c
    dont_appraise fsmagic=0xf97cff8c
    # SMACK_MAGIC=0x43415d53
    dont_appraise fsmagic=0x43415d53
    # NSFS_MAGIC=0x6e736673
    dont_appraise fsmagic=0x6e736673
    # EFIVARFS_MAGIC
    dont_appraise fsmagic=0xde5e81e4
    # CGROUP_SUPER_MAGIC=0x27e0eb
    dont_appraise fsmagic=0x27e0eb
    # CGROUP2_SUPER_MAGIC=0x63677270
    dont_appraise fsmagic=0x63677270
    appraise func=BPRM_CHECK
    appraise func=FILE_MMAP mask=MAY_EXEC
  4. To load the IMA policy to make sure the kernel accepts this IMA policy, run:

    # echo /etc/sysconfig/ima-policy > /sys/kernel/security/ima/policy
    # echo $?
    0
  5. To enable the dracut integrity module to automatically load the IMA code signing key and the IMA policy, run:

    # echo 'add_dracutmodules+=" integrity "' > /etc/dracut.conf.d/98-integrity.conf
    # dracut -f

21.9. Creating custom IMA keys using OpenSSL

You can use OpenSSL to generate a CSR for your digital certificates to secure your code.

The kernel searches the .ima keyring for a code signing key to verify an IMA signature. Before you add a code signing key to the .ima keyring, you need to ensure that IMA CA key signed this key in the .builtin_trusted_keys or .secondary_trusted_keys keyrings.

Prerequisites

  • The custom IMA CA key has the following extensions:

    • the basic constraints extension with the CA boolean asserted.
    • the KeyUsage extension with the keyCertSign bit asserted but without the digitalSignature asserted.
  • The custom IMA code signing key falls under the following criteria:

    • The IMA CA key signed this custom IMA code signing key.
    • The custom key includes the subjectKeyIdentifier extension.

Procedure

  1. To generate a custom IMA CA key pair, run:

    # openssl req -new -x509 -utf8 -sha256 -days 3650 -batch -config ima_ca.conf -outform DER -out custom_ima_ca.der -keyout custom_ima_ca.priv
  2. Optional: To check the content of the ima_ca.conf file, run:

    # cat ima_ca.conf
    [ req ]
    default_bits = 2048
    distinguished_name = req_distinguished_name
    prompt = no
    string_mask = utf8only
    x509_extensions = ca
    
    [ req_distinguished_name ]
    O = YOUR_ORG
    CN =  YOUR_COMMON_NAME IMA CA
    emailAddress = YOUR_EMAIL
    
    [ ca ]
    basicConstraints=critical,CA:TRUE
    subjectKeyIdentifier=hash
    authorityKeyIdentifier=keyid:always,issuer
    keyUsage=critical,keyCertSign,cRLSign
  3. To generate a private key and a certificate signing request (CSR) for the IMA code signing key, run:

    # openssl req -new -utf8 -sha256 -days 365 -batch -config ima.conf -out custom_ima.csr -keyout custom_ima.priv
  4. Optional: To check the content of the ima.conf file, run:

    # cat ima.conf
    [ req ]
    default_bits = 2048
    distinguished_name = req_distinguished_name
    prompt = no
    string_mask = utf8only
    x509_extensions = code_signing
    
    [ req_distinguished_name ]
    O = YOUR_ORG
    CN = YOUR_COMMON_NAME IMA signing key
    emailAddress = YOUR_EMAIL
    
    [ code_signing ]
    basicConstraints=critical,CA:FALSE
    keyUsage=digitalSignature
    subjectKeyIdentifier=hash
    authorityKeyIdentifier=keyid:always,issuer
  5. Use the IMA CA private key to sign the CSR to create the IMA code signing certificate:

    # openssl x509 -req -in custom_ima.csr -days 365 -extfile ima.conf -extensions code_signing -CA custom_ima_ca.der -CAkey custom_ima_ca.priv -CAcreateserial -outform DER -out ima.der

21.10. Deploying a custom signed IMA policy for UEFI systems

In the Secure Boot environment, you may want to only load a signed IMA policy signed by your custom IMA key.

Prerequisites

Procedure

  1. Enable Secure Boot.
  2. Permanently add the ima_policy=secure_boot kernel parameter.

    For instructions, see Configuring kernel parameters permanently with sysctl.

  3. Prepare your IMA policy by running the command:

    # evmctl ima_sign /etc/sysconfig/ima-policy -k <PATH_TO_YOUR_CUSTOM_IMA_KEY>
    Place your public certificate under /etc/keys/ima/ and add it to the .ima keyring
  4. Sign the policy with your custom IMA code signing key by running the command:

    # keyctl padd asymmetric CUSTOM_IMA1 %:.ima < /etc/ima/keys/my_ima.cer
  5. Load the IMA policy by running the command:

    # echo /etc/sysconfig/ima-policy > /sys/kernel/security/ima/policy
    # echo $?
    0

Chapter 22. Using systemd to manage resources used by applications

RHEL 9 moves the resource management settings from the process level to the application level by binding the system of cgroup hierarchies with the systemd unit tree. Therefore, you can manage the system resources with the systemctl command, or by modifying the systemd unit files.

To achieve this, systemd takes various configuration options from the unit files or directly via the systemctl command. Then systemd applies those options to specific process groups by using the Linux kernel system calls and features like cgroups and namespaces.

Note

You can review the full set of configuration options for systemd in the following manual pages:

  • systemd.resource-control(5)
  • systemd.exec(5)

22.1. Role of systemd in resource management

The core function of systemd is service management and supervision. The systemd system and service manager :

  • ensures that managed services start at the right time and in the correct order during the boot process.
  • ensures that managed services run smoothly to use the underlying hardware platform optimally.
  • provides capabilities to define resource management policies.
  • provides capabilities to tune various options, which can improve the performance of the service.
Important

In general, Red Hat recommends you use systemd for controlling the usage of system resources. You should manually configure the cgroups virtual file system only in special cases. For example, when you need to use cgroup-v1 controllers that have no equivalents in cgroup-v2 hierarchy.

22.2. Distribution models of system sources

To modify the distribution of system resources, you can apply one or more of the following distribution models:

Weights

You can distribute the resource by adding up the weights of all sub-groups and giving each sub-group the fraction matching its ratio against the sum.

For example, if you have 10 cgroups, each with weight of value 100, the sum is 1000. Each cgroup receives one tenth of the resource.

Weight is usually used to distribute stateless resources. For example the CPUWeight= option is an implementation of this resource distribution model.

Limits

A cgroup can consume up to the configured amount of the resource. The sum of sub-group limits can exceed the limit of the parent cgroup. Therefore it is possible to overcommit resources in this model.

For example the MemoryMax= option is an implementation of this resource distribution model.

Protections

You can set up a protected amount of a resource for a cgroup. If the resource usage is below the protection boundary, the kernel will try not to penalize this cgroup in favor of other cgroups that compete for the same resource. An overcommit is also possible.

For example the MemoryLow= option is an implementation of this resource distribution model.

Allocations
Exclusive allocations of an absolute amount of a finite resource. An overcommit is not possible. An example of this resource type in Linux is the real-time budget.
unit file option

A setting for resource control configuration.

For example, you can configure CPU resource with options like CPUAccounting=, or CPUQuota=. Similarly, you can configure memory or I/O resources with options like AllowedMemoryNodes= and IOAccounting=.

22.3. Allocating system resources using systemd

Procedure

To change the required value of the unit file option of your service, you can adjust the value in the unit file, or use the systemctl command:

  1. Check the assigned values for the service of your choice.

    # systemctl show --property <unit file option> <service name>
  2. Set the required value of the CPU time allocation policy option:

    # systemctl set-property <service name> <unit file option>=<value>

Verification steps

  • Check the newly assigned values for the service of your choice.

    # systemctl show --property <unit file option> <service name>

Additional resources

  • The systemd.resource-control(5) and systemd.exec(5) man pages

22.4. Overview of systemd hierarchy for cgroups

On the backend, the systemd system and service manager uses the slice, the scope, and the service units to organize and structure processes in the control groups. You can further modify this hierarchy by creating custom unit files or using the systemctl command. Also, systemd automatically mounts hierarchies for important kernel resource controllers at the /sys/fs/cgroup/ directory.

For resource control, you can use the following three systemd unit types:

Service

A process or a group of processes, which systemd started according to a unit configuration file.

Services encapsulate the specified processes so that they can be started and stopped as one set. Services are named in the following way:

<name>.service
Scope

A group of externally created processes. Scopes encapsulate processes that are started and stopped by the arbitrary processes through the fork() function and then registered by systemd at runtime. For example, user sessions, containers, and virtual machines are treated as scopes. Scopes are named as follows:

<name>.scope
Slice

A group of hierarchically organized units. Slices organize a hierarchy in which scopes and services are placed.

The actual processes are contained in scopes or in services. Every name of a slice unit corresponds to the path to a location in the hierarchy.

The dash (-) character acts as a separator of the path components to a slice from the -.slice root slice. In the following example:

<parent-name>.slice

parent-name.slice is a sub-slice of parent.slice, which is a sub-slice of the -.slice root slice. parent-name.slice can have its own sub-slice named parent-name-name2.slice, and so on.

The service, the scope, and the slice units directly map to objects in the control group hierarchy. When these units are activated, they map directly to control group paths built from the unit names.

The following is an abbreviated example of a control group hierarchy:

Control group /:
-.slice
├─user.slice
│ ├─user-42.slice
│ │ ├─session-c1.scope
│ │ │ ├─ 967 gdm-session-worker [pam/gdm-launch-environment]
│ │ │ ├─1035 /usr/libexec/gdm-x-session gnome-session --autostart /usr/share/gdm/greeter/autostart
│ │ │ ├─1054 /usr/libexec/Xorg vt1 -displayfd 3 -auth /run/user/42/gdm/Xauthority -background none -noreset -keeptty -verbose 3
│ │ │ ├─1212 /usr/libexec/gnome-session-binary --autostart /usr/share/gdm/greeter/autostart
│ │ │ ├─1369 /usr/bin/gnome-shell
│ │ │ ├─1732 ibus-daemon --xim --panel disable
│ │ │ ├─1752 /usr/libexec/ibus-dconf
│ │ │ ├─1762 /usr/libexec/ibus-x11 --kill-daemon
│ │ │ ├─1912 /usr/libexec/gsd-xsettings
│ │ │ ├─1917 /usr/libexec/gsd-a11y-settings
│ │ │ ├─1920 /usr/libexec/gsd-clipboard
…​
├─init.scope
│ └─1 /usr/lib/systemd/systemd --switched-root --system --deserialize 18
└─system.slice
  ├─rngd.service
  │ └─800 /sbin/rngd -f
  ├─systemd-udevd.service
  │ └─659 /usr/lib/systemd/systemd-udevd
  ├─chronyd.service
  │ └─823 /usr/sbin/chronyd
  ├─auditd.service
  │ ├─761 /sbin/auditd
  │ └─763 /usr/sbin/sedispatch
  ├─accounts-daemon.service
  │ └─876 /usr/libexec/accounts-daemon
  ├─example.service
  │ ├─ 929 /bin/bash /home/jdoe/example.sh
  │ └─4902 sleep 1
  …​

The example above shows that services and scopes contain processes and are placed in slices that do not contain processes of their own.

Additional resources

22.5. Listing systemd units

Use the systemd system and service manager to list its units.

Procedure

  • List all active units on the system with the systemctl utility. The terminal returns an output similar to the following example:

    # systemctl
    UNIT                                                LOAD   ACTIVE SUB       DESCRIPTION
    …​
    init.scope                                          loaded active running   System and Service Manager
    session-2.scope                                     loaded active running   Session 2 of user jdoe
    abrt-ccpp.service                                   loaded active exited    Install ABRT coredump hook
    abrt-oops.service                                   loaded active running   ABRT kernel log watcher
    abrt-vmcore.service                                 loaded active exited    Harvest vmcores for ABRT
    abrt-xorg.service                                   loaded active running   ABRT Xorg log watcher
    …​
    -.slice                                             loaded active active    Root Slice
    machine.slice                                       loaded active active    Virtual Machine and Container Slice system-getty.slice                                                                       loaded active active    system-getty.slice
    system-lvm2\x2dpvscan.slice                         loaded active active    system-lvm2\x2dpvscan.slice
    system-sshd\x2dkeygen.slice                         loaded active active    system-sshd\x2dkeygen.slice
    system-systemd\x2dhibernate\x2dresume.slice         loaded active active    system-systemd\x2dhibernate\x2dresume>
    system-user\x2druntime\x2ddir.slice                 loaded active active    system-user\x2druntime\x2ddir.slice
    system.slice                                        loaded active active    System Slice
    user-1000.slice                                     loaded active active    User Slice of UID 1000
    user-42.slice                                       loaded active active    User Slice of UID 42
    user.slice                                          loaded active active    User and Session Slice
    …​
    UNIT
    A name of a unit that also reflects the unit position in a control group hierarchy. The units relevant for resource control are a slice, a scope, and a service.
    LOAD
    Indicates whether the unit configuration file was properly loaded. If the unit file failed to load, the field contains the state error instead of loaded. Other unit load states are: stub, merged, and masked.
    ACTIVE
    The high-level unit activation state, which is a generalization of SUB.
    SUB
    The low-level unit activation state. The range of possible values depends on the unit type.
    DESCRIPTION
    The description of the unit content and functionality.
  • List all active and inactive units:

    # systemctl --all
  • Limit the amount of information in the output:

    # systemctl --type service,masked

    The --type option requires a comma-separated list of unit types such as a service and a slice, or unit load states such as loaded and masked.

Additional resources

22.6. Viewing systemd cgroups hierarchy

Display control groups (cgroups) hierarchy and processes running in specific cgroups.

Procedure

  • Display the whole cgroups hierarchy on your system with the systemd-cgls command.

    # systemd-cgls
    Control group /:
    -.slice
    ├─user.slice
    │ ├─user-42.slice
    │ │ ├─session-c1.scope
    │ │ │ ├─ 965 gdm-session-worker [pam/gdm-launch-environment]
    │ │ │ ├─1040 /usr/libexec/gdm-x-session gnome-session --autostart /usr/share/gdm/greeter/autostart
    …​
    ├─init.scope
    │ └─1 /usr/lib/systemd/systemd --switched-root --system --deserialize 18
    └─system.slice
      …​
      ├─example.service
      │ ├─6882 /bin/bash /home/jdoe/example.sh
      │ └─6902 sleep 1
      ├─systemd-journald.service
        └─629 /usr/lib/systemd/systemd-journald
      …​

    The example output returns the entire cgroups hierarchy, where the highest level is formed by slices.

  • Display the cgroups hierarchy filtered by a resource controller with the systemd-cgls <resource_controller> command.

    # systemd-cgls memory
    Controller memory; Control group /:
    ├─1 /usr/lib/systemd/systemd --switched-root --system --deserialize 18
    ├─user.slice
    │ ├─user-42.slice
    │ │ ├─session-c1.scope
    │ │ │ ├─ 965 gdm-session-worker [pam/gdm-launch-environment]
    …​
    └─system.slice
      |
      …​
      ├─chronyd.service
      │ └─844 /usr/sbin/chronyd
      ├─example.service
      │ ├─8914 /bin/bash /home/jdoe/example.sh
      │ └─8916 sleep 1
      …​

    The example output lists the services that interact with the selected controller.

  • Display detailed information about a certain unit and its part of the cgroups hierarchy with the systemctl status <system_unit> command.

    # systemctl status example.service
    ● example.service - My example service
       Loaded: loaded (/usr/lib/systemd/system/example.service; enabled; vendor preset: disabled)
       Active: active (running) since Tue 2019-04-16 12:12:39 CEST; 3s ago
     Main PID: 17737 (bash)
        Tasks: 2 (limit: 11522)
       Memory: 496.0K (limit: 1.5M)
       CGroup: /system.slice/example.service
               ├─17737 /bin/bash /home/jdoe/example.sh
               └─17743 sleep 1
    Apr 16 12:12:39 redhat systemd[1]: Started My example service.
    Apr 16 12:12:39 redhat bash[17737]: The current time is Tue Apr 16 12:12:39 CEST 2019
    Apr 16 12:12:40 redhat bash[17737]: The current time is Tue Apr 16 12:12:40 CEST 2019

Additional resources

22.7. Viewing cgroups of processes

You can learn which control group (cgroup) a process belongs to. Then you can check the cgroup to find which controllers and controller-specific configurations it uses.

Procedure

  1. To view which cgroup a process belongs to, run the # cat proc/<PID>/cgroup command:

    # cat /proc/2467/cgroup
    0::/system.slice/example.service

    The example output relates to a process of interest. In this case, it is a process identified by PID 2467, which belongs to the example.service unit. You can determine whether the process was placed in a correct control group as defined by the systemd unit file specifications.

  2. To display what controllers and respective configuration files the cgroup uses, check the cgroup directory:

    # cat /sys/fs/cgroup/system.slice/example.service/cgroup.controllers
    memory pids
    
    # ls /sys/fs/cgroup/system.slice/example.service/
    cgroup.controllers
    cgroup.events
    …​
    cpu.pressure
    cpu.stat
    io.pressure
    memory.current
    memory.events
    …​
    pids.current
    pids.events
    pids.max
Note

The version 1 hierarchy of cgroups uses a per-controller model. Therefore the output from the /proc/PID/cgroup file shows, which cgroups under each controller the PID belongs to. You can find the respective cgroups under the controller directories at /sys/fs/cgroup/<controller_name>/.

Additional resources

  • The cgroups(7) manual page
  • What are kernel resource controllers
  • Documentation in the /usr/share/doc/kernel-doc-<kernel_version>/Documentation/admin-guide/cgroup-v2.rst file (after installing the kernel-doc package)

22.8. Monitoring resource consumption

View a list of currently running control groups (cgroups) and their resource consumption in real-time.

Procedure

  1. Display a dynamic account of currently running cgroups with the systemd-cgtop command.

    # systemd-cgtop
    Control Group                            Tasks   %CPU   Memory  Input/s Output/s
    /                                          607   29.8     1.5G        -        -
    /system.slice                              125      -   428.7M        -        -
    /system.slice/ModemManager.service           3      -     8.6M        -        -
    /system.slice/NetworkManager.service         3      -    12.8M        -        -
    /system.slice/accounts-daemon.service        3      -     1.8M        -        -
    /system.slice/boot.mount                     -      -    48.0K        -        -
    /system.slice/chronyd.service                1      -     2.0M        -        -
    /system.slice/cockpit.socket                 -      -     1.3M        -        -
    /system.slice/colord.service                 3      -     3.5M        -        -
    /system.slice/crond.service                  1      -     1.8M        -        -
    /system.slice/cups.service                   1      -     3.1M        -        -
    /system.slice/dev-hugepages.mount            -      -   244.0K        -        -
    /system.slice/dev-mapper-rhel\x2dswap.swap   -      -   912.0K        -        -
    /system.slice/dev-mqueue.mount               -      -    48.0K        -        -
    /system.slice/example.service                2      -     2.0M        -        -
    /system.slice/firewalld.service              2      -    28.8M        -        -
    ...

    The example output displays currently running cgroups ordered by their resource usage (CPU, memory, disk I/O load). The list refreshes every 1 second by default. Therefore, it offers a dynamic insight into the actual resource usage of each control group.

Additional resources

  • The systemd-cgtop(1) manual page

22.9. Using systemd unit files to set limits for applications

The systemd service manager supervises each existing or running unit and creates control groups for them. The units have configuration files in the /usr/lib/systemd/system/ directory.

You can manually modify the unit files to:

  • set limits.
  • prioritize.
  • control access to hardware resources for groups of processes.

Prerequisites

  • You have the root privileges.

Procedure

  1. Edit the /usr/lib/systemd/system/example.service file to limit the memory usage of a service:

    …​
    [Service]
    MemoryMax=1500K
    …​

    The configuration limits the maximum memory that the processes in a control group cannot exceed. The example.service service is part of such a control group which has imposed limitations. You can use suffixes K, M, G, or T to identify Kilobyte, Megabyte, Gigabyte, or Terabyte as a unit of measurement.

  2. Reload all unit configuration files:

    # systemctl daemon-reload
  3. Restart the service:

    # systemctl restart example.service

Verification

  1. Check that the changes took effect:

    # cat /sys/fs/cgroup/system.slice/example.service/memory.max
    1536000

    The example output shows that the memory consumption was limited at around 1,500 KB.

Additional resources

22.10. Using systemctl command to set limits to applications

CPU affinity settings help you restrict the access of a particular process to some CPUs. Effectively, the CPU scheduler never schedules the process to run on the CPU that is not in the affinity mask of the process.

The default CPU affinity mask applies to all services managed by systemd.

To configure CPU affinity mask for a particular systemd service, systemd provides CPUAffinity= both as:

  • a unit file option.
  • a configuration option in the [Manager] section of the /etc/systemd/system.conf file.

The CPUAffinity= unit file option sets a list of CPUs or CPU ranges that are merged and used as the affinity mask.

Procedure

To set CPU affinity mask for a particular systemd service using the CPUAffinity unit file option:

  1. Check the values of the CPUAffinity unit file option in the service of your choice:

    $ systemctl show --property <CPU affinity configuration option> <service name>
  2. As the root user, set the required value of the CPUAffinity unit file option for the CPU ranges used as the affinity mask:

    # systemctl set-property <service name> CPUAffinity=<value>
  3. Restart the service to apply the changes.

    # systemctl restart <service name>

Additional resources

  • The systemd.resource-control(5), systemd.exec(5), cgroups(7) man pages

22.11. Setting global default CPU affinity through manager configuration

The CPUAffinity option in the /etc/systemd/system.conf file defines an affinity mask for the process identification number (PID) 1 and all processes forked off of PID1. You can then override the CPUAffinity on a per-service basis.

To set the default CPU affinity mask for all systemd services using the /etc/systemd/system.conf file:

  1. Set the CPU numbers for the CPUAffinity= option in the [Manager] section of the /etc/systemd/system.conf file.
  2. Save the edited file and reload the systemd service:

    # systemctl daemon-reload
  3. Reboot the server to apply the changes.

Additional resources

  • The systemd.resource-control(5) and systemd.exec(5) man pages.

22.12. Configuring NUMA policies using systemd

Non-uniform memory access (NUMA) is a computer memory subsystem design, in which the memory access time depends on the physical memory location relative to the processor.

Memory close to the CPU has lower latency (local memory) than memory that is local for a different CPU (foreign memory) or is shared between a set of CPUs.

In terms of the Linux kernel, NUMA policy governs where (for example, on which NUMA nodes) the kernel allocates physical memory pages for the process.

systemd provides unit file options NUMAPolicy and NUMAMask to control memory allocation policies for services.

Procedure

To set the NUMA memory policy through the NUMAPolicy unit file option:

  1. Check the values of the NUMAPolicy unit file option in the service of your choice:

    $ systemctl show --property <NUMA policy configuration option> <service name>
  2. As a root, set the required policy type of the NUMAPolicy unit file option:

    # systemctl set-property <service name> NUMAPolicy=<value>
  3. Restart the service to apply the changes.

    # systemctl restart <service name>

To set a global NUMAPolicy setting using the [Manager] configuration option:

  1. Search in the /etc/systemd/system.conf file for the NUMAPolicy option in the [Manager] section of the file.
  2. Edit the policy type and save the file.
  3. Reload the systemd configuration:

    # systemd daemon-reload
  4. Reboot the server.
Important

When you configure a strict NUMA policy, for example bind, make sure that you also appropriately set the CPUAffinity= unit file option.

Additional resources

22.13. NUMA policy configuration options for systemd

Systemd provides the following options to configure the NUMA policy:

NUMAPolicy

Controls the NUMA memory policy of the executed processes. You can use these policy types:

  • default
  • preferred
  • bind
  • interleave
  • local
NUMAMask

Controls the NUMA node list which is associated with the selected NUMA policy.

Note that you do not have to specify the NUMAMask option for the following policies:

  • default
  • local

For the preferred policy, the list specifies only a single NUMA node.

Additional resources

  • The systemd.resource-control(5), systemd.exec(5), and set_mempolicy(2) man pages

22.14. Creating transient cgroups using systemd-run command

The transient cgroups set limits on resources consumed by a unit (service or scope) during its runtime.

Procedure

  • To create a transient control group, use the systemd-run command in the following format:

    # systemd-run --unit=<name> --slice=<name>.slice <command>

    This command creates and starts a transient service or a scope unit and runs a custom command in such a unit.

    • The --unit=<name> option gives a name to the unit. If --unit is not specified, the name is generated automatically.
    • The --slice=<name>.slice option makes your service or scope unit a member of a specified slice. Replace <name>.slice with the name of an existing slice (as shown in the output of systemctl -t slice), or create a new slice by passing a unique name. By default, services and scopes are created as members of the system.slice.
    • Replace <command> with the command you want to enter in the service or the scope unit.

      The following message is displayed to confirm that you created and started the service or the scope successfully:

      # Running as unit <name>.service
  • Optional: Keep the unit running after its processes finished to collect run-time information:

    # systemd-run --unit=<name> --slice=<name>.slice --remain-after-exit <command>

    The command creates and starts a transient service unit and runs a custom command in the unit. The --remain-after-exit option ensures that the service keeps running after its processes have finished.

Additional resources

22.15. Removing transient control groups

You can use the systemd system and service manager to remove transient control groups (cgroups) if you no longer need to limit, prioritize, or control access to hardware resources for groups of processes.

Transient cgroups are automatically released once all the processes that a service or a scope unit contains finish.

Procedure

  • To stop the service unit with all its processes, enter:

    # systemctl stop <name>.service
  • To terminate one or more of the unit processes, enter:

    # systemctl kill <name>.service --kill-who=PID,…​ --signal=<signal>

    The command uses the --kill-who option to select process(es) from the control group you want to terminate. To kill multiple processes at the same time, pass a comma-separated list of PIDs. The --signal option determines the type of POSIX signal to be sent to the specified processes. The default signal is SIGTERM.

Additional resources

Chapter 23. Understanding control groups

Using the control groups (cgroups) kernel functionality, you can control resource usage of applications to use them more efficiently.

You can use cgroups for the following tasks:

  • Setting limits for system resource allocation.
  • Prioritizing the allocation of hardware resources to specific processes.
  • Isolating certain processes from obtaining hardware resources.

23.1. Introducing control groups

Using the control groups Linux kernel feature, you can organize processes into hierarchically ordered groups - cgroups. You define the hierarchy (control groups tree) by providing structure to cgroups virtual file system, mounted by default on the /sys/fs/cgroup/ directory.

The systemd service manager uses cgroups to organize all units and services that it governs. Manually, you can manage the hierarchies of cgroups by creating and removing sub-directories in the /sys/fs/cgroup/ directory.

The resource controllers in the kernel then modify the behavior of processes in cgroups by limiting, prioritizing or allocating system resources, of those processes. These resources include the following:

  • CPU time
  • Memory
  • Network bandwidth
  • Combinations of these resources

The primary use case of cgroups is aggregating system processes and dividing hardware resources among applications and users. This makes it possible to increase the efficiency, stability, and security of your environment.

Control groups version 1

Control groups version 1 (cgroups-v1) provide a per-resource controller hierarchy. This means that each resource (such as CPU, memory, or I/O) has its own control group hierarchy. You can combine different control group hierarchies in a way that one controller can coordinate with another in managing their respective resources. However, when the two controllers belong to different process hierarchies, proper coordination is limited.

The cgroups-v1 controllers were developed across a large time span and as a result, the behavior and naming of their control files is not uniform.

Control groups version 2

Control groups version 2 (cgroups-v2) provide a single control group hierarchy against which all resource controllers are mounted.

The control file behavior and naming is consistent among different controllers.

Important

RHEL 9, by default, mounts and uses cgroups-v2.

Additional resources

23.2. Introducing kernel resource controllers

Kernel resource controllers enable the functionality of control groups. RHEL 9 supports various controllers for control groups version 1 (cgroups-v1) and control groups version 2 (cgroups-v2).

A resource controller, also called a control group subsystem, is a kernel subsystem that represents a single resource, such as CPU time, memory, network bandwidth or disk I/O. The Linux kernel provides a range of resource controllers that are mounted automatically by the systemd service manager. You can find a list of the currently mounted resource controllers in the /proc/cgroups file.

Controllers available for cgroups-v1:

blkio
Sets limits on input/output access to and from block devices.
cpu
Adjusts the parameters of the Completely Fair Scheduler (CFS) for a control group’s tasks. The cpu controller is mounted together with the cpuacct controller on the same mount.
cpuacct
Creates automatic reports on CPU resources used by tasks in a control group. The cpuacct controller is mounted together with the cpu controller on the same mount.
cpuset
Restricts control group tasks to run only on a specified subset of CPUs and to direct the tasks to use memory only on specified memory nodes.
devices
Controls access to devices for tasks in a control group.
freezer
Suspends or resumes tasks in a control group.
memory
Sets limits on memory use by tasks in a control group and generates automatic reports on memory resources used by those tasks.
net_cls
Tags network packets with a class identifier (classid) that enables the Linux traffic controller (the tc command) to identify packets that originate from a particular control group task. A subsystem of net_cls, the net_filter (iptables), can also use this tag to perform actions on such packets. The net_filter tags network sockets with a firewall identifier (fwid) that allows the Linux firewall to identify packets that originate from a particular control group task (by using the iptables command).
net_prio
Sets the priority of network traffic.
pids
Sets limits for a number of processes and their children in a control group.
perf_event
Groups tasks for monitoring by the perf performance monitoring and reporting utility.
rdma
Sets limits on Remote Direct Memory Access/InfiniBand specific resources in a control group.
hugetlb
Can be used to limit the usage of large size virtual memory pages by tasks in a control group.

Controllers available for cgroups-v2:

io
Sets limits on input/output access to and from block devices.
memory
Sets limits on memory use by tasks in a control group and generates automatic reports on memory resources used by those tasks.
pids
Sets limits for a number of processes and their children in a control group.
rdma
Sets limits on Remote Direct Memory Access/InfiniBand specific resources in a control group.
cpu
Adjusts the parameters of the Completely Fair Scheduler (CFS) for a control group’s tasks and creates automatic reports on CPU resources used by tasks in a control group.
cpuset
Restricts control group tasks to run only on a specified subset of CPUs and to direct the tasks to use memory only on specified memory nodes. Supports only the core functionality (cpus{,.effective}, mems{,.effective}) with a new partition feature.
perf_event
Groups tasks for monitoring by the perf performance monitoring and reporting utility. perf_event is enabled automatically on the v2 hierarchy.
Important

A resource controller can be used either in a cgroups-v1 hierarchy or a cgroups-v2 hierarchy, not simultaneously in both.

Additional resources

  • The cgroups(7) manual page
  • Documentation in /usr/share/doc/kernel-doc-<kernel_version>/Documentation/cgroups-v1/ directory (after installing the kernel-doc package).

23.3. Introducing namespaces

Namespaces are one of the most important methods for organizing and identifying software objects.

A namespace wraps a global system resource (for example, a mount point, a network device, or a hostname) in an abstraction that makes it appear to processes within the namespace that they have their own isolated instance of the global resource. One of the most common technologies that use namespaces are containers.

Changes to a particular global resource are visible only to processes in that namespace and do not affect the rest of the system or other namespaces.

To inspect which namespaces a process is a member of, you can check the symbolic links in the /proc/<PID>/ns/ directory.

Table 23.1. Supported namespaces and resources which they isolate:

NamespaceIsolates

Mount

Mount points

UTS

Hostname and NIS domain name

IPC

System V IPC, POSIX message queues

PID

Process IDs

Network

Network devices, stacks, ports, etc

User

User and group IDs

Control groups

Control group root directory

Additional resources

Chapter 24. Using cgroupfs to manually manage cgroups

You can manage cgroup hierarchies on your system by creating directories on the cgroupfs virtual file system. The file system is mounted by default on the /sys/fs/cgroup/ directory and you can specify desired configurations in dedicated control files.

Important

In general, Red Hat recommends you use systemd for controlling the usage of system resources. You should manually configure the cgroups virtual file system only in special cases. For example, when you need to use cgroup-v1 controllers that have no equivalents in cgroup-v2 hierarchy.

24.1. Creating cgroups and enabling controllers in cgroups-v2 file system

You can manage the control groups (cgroups) by creating or removing directories and by writing to files in the cgroups virtual file system. The file system is by default mounted on the /sys/fs/cgroup/ directory. To use settings from the cgroups controllers, you also need to enable the desired controllers for child cgroups. The root cgroup has, by default, enabled the memory and pids controllers for its child cgroups. Therefore, Red Hat recommends to create at least two levels of child cgroups inside the /sys/fs/cgroup/ root cgroup. This way you optionally remove the memory and pids controllers from the child cgroups and maintain better organizational clarity of cgroup files.

Prerequisites

  • You have root permissions.

Procedure

  1. Create the /sys/fs/cgroup/Example/ directory:

    # mkdir /sys/fs/cgroup/Example/

    The /sys/fs/cgroup/Example/ directory defines a child group. When you create the /sys/fs/cgroup/Example/ directory, some cgroups-v2 interface files are automatically created in the directory. The /sys/fs/cgroup/Example/ directory contains also controller-specific files for the memory and pids controllers.

  2. Optionally, inspect the newly created child control group:

    # ll /sys/fs/cgroup/Example/
    -r—​r—​r--. 1 root root 0 Jun  1 10:33 cgroup.controllers
    -r—​r—​r--. 1 root root 0 Jun  1 10:33 cgroup.events
    -rw-r—​r--. 1 root root 0 Jun  1 10:33 cgroup.freeze
    -rw-r—​r--. 1 root root 0 Jun  1 10:33 cgroup.procs
    …​
    -rw-r—​r--. 1 root root 0 Jun  1 10:33 cgroup.subtree_control
    -r—​r—​r--. 1 root root 0 Jun  1 10:33 memory.events.local
    -rw-r—​r--. 1 root root 0 Jun  1 10:33 memory.high
    -rw-r—​r--. 1 root root 0 Jun  1 10:33 memory.low
    …​
    -r—​r—​r--. 1 root root 0 Jun  1 10:33 pids.current
    -r—​r—​r--. 1 root root 0 Jun  1 10:33 pids.events
    -rw-r—​r--. 1 root root 0 Jun  1 10:33 pids.max

    The example output shows general cgroup control interface files such as cgroup.procs or cgroup.controllers. These files are common to all control groups, regardless of enabled controllers.

    The files such as memory.high and pids.max relate to the memory and pids controllers, which are in the root control group (/sys/fs/cgroup/), and are enabled by default by systemd.

    By default, the newly created child group inherits all settings from the parent cgroup. In this case, there are no limits from the root cgroup.

  3. Verify that the desired controllers are available in the /sys/fs/cgroup/cgroup.controllers file:

    # cat /sys/fs/cgroup/cgroup.controllers
    cpuset cpu io memory hugetlb pids rdma
  4. Enable the desired controllers. In this example it is cpu and cpuset controllers:

    # echo "+cpu" >> /sys/fs/cgroup/cgroup.subtree_control
    # echo "+cpuset" >> /sys/fs/cgroup/cgroup.subtree_control

    These commands enable the cpu and cpuset controllers for the immediate child groups of the /sys/fs/cgroup/ root control group. Including the newly created Example control group. A child group is where you can specify processes and apply control checks to each of the processes based on your criteria.

    Users can read the contents of the cgroup.subtree_control file at any level to get an idea of what controllers are going to be available for enablement in the immediate child group.

    Note

    By default, the /sys/fs/cgroup/cgroup.subtree_control file in the root control group contains memory and pids controllers.

  5. Enable the desired controllers for child cgroups of the Example control group:

    # echo "+cpu +cpuset" >> /sys/fs/cgroup/Example/cgroup.subtree_control

    This command ensures that the immediate child control group will only have controllers relevant to regulate the CPU time distribution - not to memory or pids controllers.

  6. Create the /sys/fs/cgroup/Example/tasks/ directory:

    # mkdir /sys/fs/cgroup/Example/tasks/

    The /sys/fs/cgroup/Example/tasks/ directory defines a child group with files that relate purely to cpu and cpuset controllers. You can now assign processes to this control group and utilize cpu and cpuset controller options for your processes.

  7. Optionally, inspect the child control group:

    # ll /sys/fs/cgroup/Example/tasks
    -r—​r—​r--. 1 root root 0 Jun  1 11:45 cgroup.controllers
    -r—​r—​r--. 1 root root 0 Jun  1 11:45 cgroup.events
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 cgroup.freeze
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 cgroup.max.depth
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 cgroup.max.descendants
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 cgroup.procs
    -r—​r—​r--. 1 root root 0 Jun  1 11:45 cgroup.stat
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 cgroup.subtree_control
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 cgroup.threads
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 cgroup.type
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 cpu.max
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 cpu.pressure
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 cpuset.cpus
    -r—​r—​r--. 1 root root 0 Jun  1 11:45 cpuset.cpus.effective
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 cpuset.cpus.partition
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 cpuset.mems
    -r—​r—​r--. 1 root root 0 Jun  1 11:45 cpuset.mems.effective
    -r—​r—​r--. 1 root root 0 Jun  1 11:45 cpu.stat
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 cpu.weight
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 cpu.weight.nice
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 io.pressure
    -rw-r—​r--. 1 root root 0 Jun  1 11:45 memory.pressure
Important

The cpu controller is only activated if the relevant child control group has at least 2 processes which compete for time on a single CPU.

Verification steps

  • Optional: confirm that you have created a new cgroup with only the desired controllers active:

    # cat /sys/fs/cgroup/Example/tasks/cgroup.controllers
    cpuset cpu

Additional resources

24.2. Controlling distribution of CPU time for applications by adjusting CPU weight

You need to assign values to the relevant files of the cpu controller to regulate distribution of the CPU time to applications under the specific cgroup tree.

Prerequisites

  • You have root permissions.
  • You have applications for which you want to control distribution of CPU time.
  • You created a two level hierarchy of child control groups inside the /sys/fs/cgroup/ root control group as in the following example:

    …​
      ├── Example
      │   ├── g1
      │   ├── g2
      │   └── g3
    …​
  • You enabled the cpu controller in the parent control group and in child control groups similarly as described in Creating cgroups and enabling controllers in cgroups-v2 file system.

Procedure

  1. Configure desired CPU weights to achieve resource restrictions within the control groups:

    # echo "150" > /sys/fs/cgroup/Example/g1/cpu.weight
    # echo "100" > /sys/fs/cgroup/Example/g2/cpu.weight
    # echo "50" > /sys/fs/cgroup/Example/g3/cpu.weight
  2. Add the applications' PIDs to the g1, g2, and g3 child groups:

    # echo "33373" > /sys/fs/cgroup/Example/g1/cgroup.procs
    # echo "33374" > /sys/fs/cgroup/Example/g2/cgroup.procs
    # echo "33377" > /sys/fs/cgroup/Example/g3/cgroup.procs

    The example commands ensure that desired applications become members of the Example/g*/ child cgroups and will get their CPU time distributed as per the configuration of those cgroups.

    The weights of the children cgroups (g1, g2, g3) that have running processes are summed up at the level of the parent cgroup (Example). The CPU resource is then distributed proportionally based on the respective weights.

    As a result, when all processes run at the same time, the kernel allocates to each of them the proportionate CPU time based on their respective cgroup’s cpu.weight file:

    Child cgroupcpu.weight fileCPU time allocation

    g1

    150

    ~50% (150/300)

    g2

    100

    ~33% (100/300)

    g3

    50

    ~16% (50/300)

    The value of the cpu.weight controller file is not a percentage.

    If one process stopped running, leaving cgroup g2 with no running processes, the calculation would omit the cgroup g2 and only account weights of cgroups g1 and g3:

    Child cgroupcpu.weight fileCPU time allocation

    g1

    150

    ~75% (150/200)

    g3

    50

    ~25% (50/200)

    Important

    If a child cgroup had multiple running processes, the CPU time allocated to the respective cgroup would be distributed equally to the member processes of that cgroup.

Verification

  1. Verify that the applications run in the specified control groups:

    # cat /proc/33373/cgroup /proc/33374/cgroup /proc/33377/cgroup
    0::/Example/g1
    0::/Example/g2
    0::/Example/g3

    The command output shows the processes of the specified applications that run in the Example/g*/ child cgroups.

  2. Inspect the current CPU consumption of the throttled applications:

    # top
    top - 05:17:18 up 1 day, 18:25,  1 user,  load average: 3.03, 3.03, 3.00
    Tasks:  95 total,   4 running,  91 sleeping,   0 stopped,   0 zombie
    %Cpu(s): 18.1 us, 81.6 sy,  0.0 ni,  0.0 id,  0.0 wa,  0.3 hi,  0.0 si,  0.0 st
    MiB Mem :   3737.0 total,   3233.7 free,    132.8 used,    370.5 buff/cache
    MiB Swap:   4060.0 total,   4060.0 free,      0.0 used.   3373.1 avail Mem
    
        PID USER      PR  NI    VIRT    RES    SHR S  %CPU  %MEM     TIME+ COMMAND
      33373 root      20   0   18720   1748   1460 R  49.5   0.0 415:05.87 sha1sum
      33374 root      20   0   18720   1756   1464 R  32.9   0.0 412:58.33 sha1sum
      33377 root      20   0   18720   1860   1568 R  16.3   0.0 411:03.12 sha1sum
        760 root      20   0  416620  28540  15296 S   0.3   0.7   0:10.23 tuned
          1 root      20   0  186328  14108   9484 S   0.0   0.4   0:02.00 systemd
          2 root      20   0       0      0      0 S   0.0   0.0   0:00.01 kthread
    ...
    Note

    We forced all the example processes to run on a single CPU for clearer illustration. The CPU weight applies the same principles also when used on multiple CPUs.

    Notice that the CPU resource for the PID 33373, PID 33374, and PID 33377 was allocated based on the weights, 150, 100, 50, you assigned to the respective child cgroups. The weights correspond to around 50%, 33%, and 16% allocation of CPU time for each application.

24.3. Mounting cgroups-v1

During the boot process, RHEL 9 mounts the cgroup-v2 virtual filesystem by default. To utilize cgroup-v1 functionality in limiting resources for your applications, manually configure the system.

Note

Both cgroup-v1 and cgroup-v2 are fully enabled in the kernel. There is no default control group version from the kernel point of view, and is decided by systemd to mount at startup.

Prerequisites

  • You have root permissions.

Procedure

  1. Configure the system to mount cgroups-v1 by default during system boot by the systemd system and service manager:

    # grubby --update-kernel=/boot/vmlinuz-$(uname -r) --args="systemd.unified_cgroup_hierarchy=0 systemd.legacy_systemd_cgroup_controller"

    This adds the necessary kernel command-line parameters to the current boot entry.

    To add the same parameters to all kernel boot entries:

    # grubby --update-kernel=ALL --args="systemd.unified_cgroup_hierarchy=0 systemd.legacy_systemd_cgroup_controller"
  2. Reboot the system for the changes to take effect.

Verification

  1. Optionally, verify that the cgroups-v1 filesystem was mounted:

    # mount -l | grep cgroup
    tmpfs on /sys/fs/cgroup type tmpfs (ro,nosuid,nodev,noexec,seclabel,size=4096k,nr_inodes=1024,mode=755,inode64)
    cgroup on /sys/fs/cgroup/systemd type cgroup (rw,nosuid,nodev,noexec,relatime,seclabel,xattr,release_agent=/usr/lib/systemd/systemd-cgroups-agent,name=systemd)
    cgroup on /sys/fs/cgroup/perf_event type cgroup (rw,nosuid,nodev,noexec,relatime,seclabel,perf_event)
    cgroup on /sys/fs/cgroup/cpu,cpuacct type cgroup (rw,nosuid,nodev,noexec,relatime,seclabel,cpu,cpuacct)
    cgroup on /sys/fs/cgroup/pids type cgroup (rw,nosuid,nodev,noexec,relatime,seclabel,pids)
    cgroup on /sys/fs/cgroup/cpuset type cgroup (rw,nosuid,nodev,noexec,relatime,seclabel,cpuset)
    cgroup on /sys/fs/cgroup/net_cls,net_prio type cgroup (rw,nosuid,nodev,noexec,relatime,seclabel,net_cls,net_prio)
    cgroup on /sys/fs/cgroup/hugetlb type cgroup (rw,nosuid,nodev,noexec,relatime,seclabel,hugetlb)
    cgroup on /sys/fs/cgroup/memory type cgroup (rw,nosuid,nodev,noexec,relatime,seclabel,memory)
    cgroup on /sys/fs/cgroup/blkio type cgroup (rw,nosuid,nodev,noexec,relatime,seclabel,blkio)
    cgroup on /sys/fs/cgroup/devices type cgroup (rw,nosuid,nodev,noexec,relatime,seclabel,devices)
    cgroup on /sys/fs/cgroup/misc type cgroup (rw,nosuid,nodev,noexec,relatime,seclabel,misc)
    cgroup on /sys/fs/cgroup/freezer type cgroup (rw,nosuid,nodev,noexec,relatime,seclabel,freezer)
    cgroup on /sys/fs/cgroup/rdma type cgroup (rw,nosuid,nodev,noexec,relatime,seclabel,rdma)

    The cgroups-v1 filesystems that correspond to various cgroup-v1 controllers, were successfully mounted on the /sys/fs/cgroup/ directory.

  2. Optionally, inspect the contents of the /sys/fs/cgroup/ directory:

    # ll /sys/fs/cgroup/
    dr-xr-xr-x. 10 root root  0 Mar 16 09:34 blkio
    lrwxrwxrwx.  1 root root 11 Mar 16 09:34 cpu → cpu,cpuacct
    lrwxrwxrwx.  1 root root 11 Mar 16 09:34 cpuacct → cpu,cpuacct
    dr-xr-xr-x. 10 root root  0 Mar 16 09:34 cpu,cpuacct
    dr-xr-xr-x.  2 root root  0 Mar 16 09:34 cpuset
    dr-xr-xr-x. 10 root root  0 Mar 16 09:34 devices
    dr-xr-xr-x.  2 root root  0 Mar 16 09:34 freezer
    dr-xr-xr-x.  2 root root  0 Mar 16 09:34 hugetlb
    dr-xr-xr-x. 10 root root  0 Mar 16 09:34 memory
    dr-xr-xr-x.  2 root root  0 Mar 16 09:34 misc
    lrwxrwxrwx.  1 root root 16 Mar 16 09:34 net_cls → net_cls,net_prio
    dr-xr-xr-x.  2 root root  0 Mar 16 09:34 net_cls,net_prio
    lrwxrwxrwx.  1 root root 16 Mar 16 09:34 net_prio → net_cls,net_prio
    dr-xr-xr-x.  2 root root  0 Mar 16 09:34 perf_event
    dr-xr-xr-x. 10 root root  0 Mar 16 09:34 pids
    dr-xr-xr-x.  2 root root  0 Mar 16 09:34 rdma
    dr-xr-xr-x. 11 root root  0 Mar 16 09:34 systemd

    The /sys/fs/cgroup/ directory, also called the root control group, by default, contains controller-specific directories such as cpuset. In addition, there are some directories related to systemd.

24.4. Setting CPU limits to applications using cgroups-v1

To configure CPU limits to an application by using control groups version 1 (cgroups-v1), use the /sys/fs/ virtual file system.

Prerequisites

  • You have root permissions.
  • You have an application whose CPU consumption you want to restrict.
  • You configured the system to mount cgroups-v1 by default during system boot by the systemd system and service manager:

    # grubby --update-kernel=/boot/vmlinuz-$(uname -r) --args="systemd.unified_cgroup_hierarchy=0 systemd.legacy_systemd_cgroup_controller"

    This adds the necessary kernel command-line parameters to the current boot entry.

Procedure

  1. Identify the process ID (PID) of the application that you want to restrict in CPU consumption:

    # top
    top - 11:34:09 up 11 min,  1 user,  load average: 0.51, 0.27, 0.22
    Tasks: 267 total,   3 running, 264 sleeping,   0 stopped,   0 zombie
    %Cpu(s): 49.0 us,  3.3 sy,  0.0 ni, 47.5 id,  0.0 wa,  0.2 hi,  0.0 si,  0.0 st
    MiB Mem :   1826.8 total,    303.4 free,   1046.8 used,    476.5 buff/cache
    MiB Swap:   1536.0 total,   1396.0 free,    140.0 used.    616.4 avail Mem
    
      PID USER      PR  NI    VIRT    RES    SHR S  %CPU  %MEM     TIME+ COMMAND
     6955 root      20   0  228440   1752   1472 R  99.3   0.1   0:32.71 sha1sum
     5760 jdoe      20   0 3603868 205188  64196 S   3.7  11.0   0:17.19 gnome-shell
     6448 jdoe      20   0  743648  30640  19488 S   0.7   1.6   0:02.73 gnome-terminal-
        1 root      20   0  245300   6568   4116 S   0.3   0.4   0:01.87 systemd
      505 root      20   0       0      0      0 I   0.3   0.0   0:00.75 kworker/u4:4-events_unbound
    ...

    This example output of the top program reveals that illustrative application sha1sum with PID 6955 consumes a lot of CPU resources.

  2. Create a sub-directory in the cpu resource controller directory:

    # mkdir /sys/fs/cgroup/cpu/Example/

    This directory represents a control group, where you can place specific processes and apply certain CPU limits to the processes. At the same time, a number of cgroups-v1 interface files and cpu controller-specific files will be created in the directory.

  3. Optional: Inspect the newly created control group:

    # ll /sys/fs/cgroup/cpu/Example/
    -rw-r—​r--. 1 root root 0 Mar 11 11:42 cgroup.clone_children
    -rw-r—​r--. 1 root root 0 Mar 11 11:42 cgroup.procs
    -r—​r—​r--. 1 root root 0 Mar 11 11:42 cpuacct.stat
    -rw-r—​r--. 1 root root 0 Mar 11 11:42 cpuacct.usage
    -r—​r—​r--. 1 root root 0 Mar 11 11:42 cpuacct.usage_all
    -r—​r—​r--. 1 root root 0 Mar 11 11:42 cpuacct.usage_percpu
    -r—​r—​r--. 1 root root 0 Mar 11 11:42 cpuacct.usage_percpu_sys
    -r—​r—​r--. 1 root root 0 Mar 11 11:42 cpuacct.usage_percpu_user
    -r—​r—​r--. 1 root root 0 Mar 11 11:42 cpuacct.usage_sys
    -r—​r—​r--. 1 root root 0 Mar 11 11:42 cpuacct.usage_user
    -rw-r—​r--. 1 root root 0 Mar 11 11:42 cpu.cfs_period_us
    -rw-r—​r--. 1 root root 0 Mar 11 11:42 cpu.cfs_quota_us
    -rw-r—​r--. 1 root root 0 Mar 11 11:42 cpu.rt_period_us
    -rw-r—​r--. 1 root root 0 Mar 11 11:42 cpu.rt_runtime_us
    -rw-r—​r--. 1 root root 0 Mar 11 11:42 cpu.shares
    -r—​r—​r--. 1 root root 0 Mar 11 11:42 cpu.stat
    -rw-r—​r--. 1 root root 0 Mar 11 11:42 notify_on_release
    -rw-r—​r--. 1 root root 0 Mar 11 11:42 tasks

    This example output shows files, such as cpuacct.usage, cpu.cfs._period_us, that represent specific configurations and/or limits, which can be set for processes in the Example control group. Note that the respective file names are prefixed with the name of the control group controller to which they belong.

    By default, the newly created control group inherits access to the system’s entire CPU resources without a limit.

  4. Configure CPU limits for the control group:

    # echo "1000000" > /sys/fs/cgroup/cpu/Example/cpu.cfs_period_us
    # echo "200000" > /sys/fs/cgroup/cpu/Example/cpu.cfs_quota_us
    • The cpu.cfs_period_us file represents a period of time in microseconds (µs, represented here as "us") for how frequently a control group’s access to CPU resources should be reallocated. The upper limit is 1 000 000 microsecond and the lower limit is 1 000 microseconds.
    • The cpu.cfs_quota_us file represents the total amount of time in microseconds for which all processes collectively in a control group can run during one period (as defined by cpu.cfs_period_us). When processes in a control group, during a single period, use up all the time specified by the quota, they are throttled for the remainder of the period and not allowed to run until the next period. The lower limit is 1 000 microseconds.

      The example commands above set the CPU time limits so that all processes collectively in the Example control group will be able to run only for 0.2 seconds (defined by cpu.cfs_quota_us) out of every 1 second (defined by cpu.cfs_period_us).

  5. Optional: Verify the limits:

    # cat /sys/fs/cgroup/cpu/Example/cpu.cfs_period_us /sys/fs/cgroup/cpu/Example/cpu.cfs_quota_us
    1000000
    200000
  6. Add the application’s PID to the Example control group:

    # echo "6955" > /sys/fs/cgroup/cpu/Example/cgroup.procs

    This command ensures that a specific application becomes a member of the Example control group and hence does not exceed the CPU limits configured for the Example control group. The PID should represent an existing process in the system. The PID 6955 here was assigned to process sha1sum /dev/zero &, used to illustrate the use case of the cpu controller.

Verification

  1. Verify that the application runs in the specified control group:

    # cat /proc/6955/cgroup
    12:cpuset:/
    11:hugetlb:/
    10:net_cls,net_prio:/
    9:memory:/user.slice/user-1000.slice/user@1000.service
    8:devices:/user.slice
    7:blkio:/
    6:freezer:/
    5:rdma:/
    4:pids:/user.slice/user-1000.slice/user@1000.service
    3:perf_event:/
    2:cpu,cpuacct:/Example
    1:name=systemd:/user.slice/user-1000.slice/user@1000.service/gnome-terminal-server.service

    This example output shows that the process of the desired application runs in the Example control group, which applies CPU limits to the application’s process.

  2. Identify the current CPU consumption of your throttled application:

    # top
    top - 12:28:42 up  1:06,  1 user,  load average: 1.02, 1.02, 1.00
    Tasks: 266 total,   6 running, 260 sleeping,   0 stopped,   0 zombie
    %Cpu(s): 11.0 us,  1.2 sy,  0.0 ni, 87.5 id,  0.0 wa,  0.2 hi,  0.0 si,  0.2 st
    MiB Mem :   1826.8 total,    287.1 free,   1054.4 used,    485.3 buff/cache
    MiB Swap:   1536.0 total,   1396.7 free,    139.2 used.    608.3 avail Mem
    
      PID USER      PR  NI    VIRT    RES    SHR S  %CPU  %MEM     TIME+ COMMAND
     6955 root      20   0  228440   1752   1472 R  20.6   0.1  47:11.43 sha1sum
     5760 jdoe      20   0 3604956 208832  65316 R   2.3  11.2   0:43.50 gnome-shell
     6448 jdoe      20   0  743836  31736  19488 S   0.7   1.7   0:08.25 gnome-terminal-
      505 root      20   0       0      0      0 I   0.3   0.0   0:03.39 kworker/u4:4-events_unbound
     4217 root      20   0   74192   1612   1320 S   0.3   0.1   0:01.19 spice-vdagentd
    ...

    Note that the CPU consumption of the PID 6955 has decreased from 99% to 20%.

Note

The cgroups-v2 counterpart for cpu.cfs_period_us and cpu.cfs_quota_us is the cpu.max file. The cpu.max file is available through the cpu controller.

Additional resources

Chapter 25. Analyzing system performance with BPF Compiler Collection

As a system administrator, you can use the BPF Compiler Collection (BCC) library to create tools for analyzing the performance of your Linux operating system and gathering information, which could be difficult to obtain through other interfaces.

25.1. Installing the bcc-tools package

Install the bcc-tools package, which also installs the BPF Compiler Collection (BCC) library as a dependency.

Procedure

  1. Install bcc-tools.

    # dnf install bcc-tools

    The BCC tools are installed in the /usr/share/bcc/tools/ directory.

  2. Optionally, inspect the tools:

    # ll /usr/share/bcc/tools/
    ...
    -rwxr-xr-x. 1 root root  4198 Dec 14 17:53 dcsnoop
    -rwxr-xr-x. 1 root root  3931 Dec 14 17:53 dcstat
    -rwxr-xr-x. 1 root root 20040 Dec 14 17:53 deadlock_detector
    -rw-r--r--. 1 root root  7105 Dec 14 17:53 deadlock_detector.c
    drwxr-xr-x. 3 root root  8192 Mar 11 10:28 doc
    -rwxr-xr-x. 1 root root  7588 Dec 14 17:53 execsnoop
    -rwxr-xr-x. 1 root root  6373 Dec 14 17:53 ext4dist
    -rwxr-xr-x. 1 root root 10401 Dec 14 17:53 ext4slower
    ...

    The doc directory in the listing above contains documentation for each tool.

25.2. Using selected bcc-tools for performance analyses

Use certain pre-created programs from the BPF Compiler Collection (BCC) library to efficiently and securely analyze the system performance on the per-event basis. The set of pre-created programs in the BCC library can serve as examples for creation of additional programs.

Prerequisites

Using execsnoop to examine the system processes

  1. Run the execsnoop program in one terminal:

    # /usr/share/bcc/tools/execsnoop
  2. In another terminal run, for example:

    $ ls /usr/share/bcc/tools/doc/

    The above creates a short-lived process of the ls command.

  3. The terminal running execsnoop shows the output similar to the following:

    PCOMM	PID    PPID   RET ARGS
    ls   	8382   8287     0 /usr/bin/ls --color=auto /usr/share/bcc/tools/doc/
    ...

    The execsnoop program prints a line of output for each new process, which consumes system resources. It even detects processes of programs that run very shortly, such as ls, and most monitoring tools would not register them.

    The execsnoop output displays the following fields:

    PCOMM
    The parent process name. (ls)
    PID
    The process ID. (8382)
    PPID
    The parent process ID. (8287)
    RET
    The return value of the exec() system call (0), which loads program code into new processes.
    ARGS
    The location of the started program with arguments.

To see more details, examples, and options for execsnoop, refer to the /usr/share/bcc/tools/doc/execsnoop_example.txt file.

For more information about exec(), see exec(3) manual pages.

Using opensnoop to track what files a command opens

  1. Run the opensnoop program in one terminal:

    # /usr/share/bcc/tools/opensnoop -n uname

    The above prints output for files, which are opened only by the process of the uname command.

  2. In another terminal, enter:

    $ uname

    The command above opens certain files, which are captured in the next step.

  3. The terminal running opensnoop shows the output similar to the following:

    PID    COMM 	FD ERR PATH
    8596   uname 	3  0   /etc/ld.so.cache
    8596   uname 	3  0   /lib64/libc.so.6
    8596   uname 	3  0   /usr/lib/locale/locale-archive
    ...

    The opensnoop program watches the open() system call across the whole system, and prints a line of output for each file that uname tried to open along the way.

    The opensnoop output displays the following fields:

    PID
    The process ID. (8596)
    COMM
    The process name. (uname)
    FD
    The file descriptor - a value that open() returns to refer to the open file. (3)
    ERR
    Any errors.
    PATH
    The location of files that open() tried to open.

    If a command tries to read a non-existent file, then the FD column returns -1 and the ERR column prints a value corresponding to the relevant error. As a result, opensnoop can help you identify an application that does not behave properly.

To see more details, examples, and options for opensnoop, refer to the /usr/share/bcc/tools/doc/opensnoop_example.txt file.

For more information about open(), see open(2) manual pages.

Using biotop to examine the I/O operations on the disk

  1. Run the biotop program in one terminal:

    # /usr/share/bcc/tools/biotop 30

    The command enables you to monitor the top processes, which perform I/O operations on the disk. The argument ensures that the command will produce a 30 second summary.

    Note

    When no argument provided, the output screen by default refreshes every 1 second.

  2. In another terminal enter, for example :

    # dd if=/dev/vda of=/dev/zero

    The command above reads the content from the local hard disk device and writes the output to the /dev/zero file. This step generates certain I/O traffic to illustrate biotop.

  3. The terminal running biotop shows the output similar to the following:

    PID    COMM             D MAJ MIN DISK       I/O  Kbytes     AVGms
    9568   dd               R 252 0   vda      16294 14440636.0  3.69
    48     kswapd0          W 252 0   vda       1763 120696.0    1.65
    7571   gnome-shell      R 252 0   vda        834 83612.0     0.33
    1891   gnome-shell      R 252 0   vda       1379 19792.0     0.15
    7515   Xorg             R 252 0   vda        280  9940.0     0.28
    7579   llvmpipe-1       R 252 0   vda        228  6928.0     0.19
    9515   gnome-control-c  R 252 0   vda         62  6444.0     0.43
    8112   gnome-terminal-  R 252 0   vda         67  2572.0     1.54
    7807   gnome-software   R 252 0   vda         31  2336.0     0.73
    9578   awk              R 252 0   vda         17  2228.0     0.66
    7578   llvmpipe-0       R 252 0   vda        156  2204.0     0.07
    9581   pgrep            R 252 0   vda         58  1748.0     0.42
    7531   InputThread      R 252 0   vda         30  1200.0     0.48
    7504   gdbus            R 252 0   vda          3  1164.0     0.30
    1983   llvmpipe-1       R 252 0   vda         39   724.0     0.08
    1982   llvmpipe-0       R 252 0   vda         36   652.0     0.06
    ...

    The biotop output displays the following fields:

    PID
    The process ID. (9568)
    COMM
    The process name. (dd)
    DISK
    The disk performing the read operations. (vda)
    I/O
    The number of read operations performed. (16294)
    Kbytes
    The amount of Kbytes reached by the read operations. (14,440,636)
    AVGms
    The average I/O time of read operations. (3.69)

To see more details, examples, and options for biotop, refer to the /usr/share/bcc/tools/doc/biotop_example.txt file.

For more information about dd, see dd(1) manual pages.

Using xfsslower to expose unexpectedly slow file system operations

  1. Run the xfsslower program in one terminal:

    # /usr/share/bcc/tools/xfsslower 1

    The command above measures the time the XFS file system spends in performing read, write, open or sync (fsync) operations. The 1 argument ensures that the program shows only the operations that are slower than 1 ms.

    Note

    When no arguments provided, xfsslower by default displays operations slower than 10 ms.

  2. In another terminal enter, for example, the following:

    $ vim text

    The command above creates a text file in the vim editor to initiate certain interaction with the XFS file system.

  3. The terminal running xfsslower shows something similar upon saving the file from the previous step:

    TIME     COMM           PID    T BYTES   OFF_KB   LAT(ms) FILENAME
    13:07:14 b'bash'        4754   R 256     0           7.11 b'vim'
    13:07:14 b'vim'         4754   R 832     0           4.03 b'libgpm.so.2.1.0'
    13:07:14 b'vim'         4754   R 32      20          1.04 b'libgpm.so.2.1.0'
    13:07:14 b'vim'         4754   R 1982    0           2.30 b'vimrc'
    13:07:14 b'vim'         4754   R 1393    0           2.52 b'getscriptPlugin.vim'
    13:07:45 b'vim'         4754   S 0       0           6.71 b'text'
    13:07:45 b'pool'        2588   R 16      0           5.58 b'text'
    ...

    Each line above represents an operation in the file system, which took more time than a certain threshold. xfsslower is good at exposing possible file system problems, which can take form of unexpectedly slow operations.

    The xfsslower output displays the following fields:

    COMM
    The process name. (b’bash')
    T

    The operation type. (R)

    • Read
    • Write
    • Sync
    OFF_KB
    The file offset in KB. (0)
    FILENAME
    The file being read, written, or synced.

To see more details, examples, and options for xfsslower, refer to the /usr/share/bcc/tools/doc/xfsslower_example.txt file.

For more information about fsync, see fsync(2) manual pages.

Legal Notice

Copyright © 2024 Red Hat, Inc.
The text of and illustrations in this document are licensed by Red Hat under a Creative Commons Attribution–Share Alike 3.0 Unported license ("CC-BY-SA"). An explanation of CC-BY-SA is available at http://creativecommons.org/licenses/by-sa/3.0/. In accordance with CC-BY-SA, if you distribute this document or an adaptation of it, you must provide the URL for the original version.
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, the Red Hat 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 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.