From Fedora Project Wiki

No edit summary
No edit summary
 
(2 intermediate revisions by the same user not shown)
Line 6: Line 6:


|actions=
|actions=
# [https://docs.fedoraproject.org/en-US/fedora-coreos/producing-ign/ Create an Ignition file]. When writing the configuration, add a `crashkernel=300M` section following the example in [https://docs.fedoraproject.org/en-US/fedora-coreos/kernel-args/#_modifying_kernel_arguments_via_ignition the documentation]. You don't necessarily have to use the same kernel arguments from the documentation, but you can.
# [https://docs.fedoraproject.org/en-US/fedora-coreos/producing-ign/ Create an Ignition file]. When writing the configuration, add a `crashkernel=300M` section following the example in [https://docs.fedoraproject.org/en-US/fedora-coreos/kernel-args/#_modifying_kernel_arguments_via_ignition the documentation]. You don't necessarily need to add it in the ignition, you can do it via rpm-ostree as described in this documentation, you also need to make sure the kdump.service will be enabled, you can do same via ignition or via `systemctl enable kdump.service`.
# Launch your Fedora CoreOS system using this Ignition file. See [https://docs.fedoraproject.org/en-US/fedora-coreos/getting-started/#_quickstart the Quickstart] and the provisioning docs for a guide how to launch FCOS in most environments (bare metal, virtual, cloud, etc).
# Launch your Fedora CoreOS system using this Ignition file. See [https://docs.fedoraproject.org/en-US/fedora-coreos/getting-started/#_quickstart the Quickstart] and the provisioning docs for a guide how to launch FCOS in most environments (bare metal, virtual, cloud, etc).
# Acess your Fedora CoreOS and force a kernel panic for testing.
  sudo systemctl enable kdump.service
  echo 1 > /proc/sys/kernel/sysrq
  echo "c" > /proc/sysrq-trigger


|results=
|results=
 
# You can SSH successfully into your FCOS system.
 
# The kernel arguments (as seen in the /proc/cmdline file) are as specified in the Ignition config, or via rpm-ostree after the reboot
# After FCOS is rebooted you can successfully see the vmcore and other log files created under `/var/crash`
# The linked documentation makes sense, nothing is obviously missing, and the instructions are clear.
}}
}}
[[Category:CoreOS Test Cases]]
[[Category:CoreOS Test Cases]]

Latest revision as of 16:50, 6 April 2022

Description

Install Fedora CoreOS and debug kernel crashes using kdump as described in documentation.

Setup

  1. Download and verify the latest FCOS next image (pick the right image for your environment).

How to test

  1. Create an Ignition file. When writing the configuration, add a crashkernel=300M section following the example in the documentation. You don't necessarily need to add it in the ignition, you can do it via rpm-ostree as described in this documentation, you also need to make sure the kdump.service will be enabled, you can do same via ignition or via systemctl enable kdump.service.
  2. Launch your Fedora CoreOS system using this Ignition file. See the Quickstart and the provisioning docs for a guide how to launch FCOS in most environments (bare metal, virtual, cloud, etc).
  3. Acess your Fedora CoreOS and force a kernel panic for testing.
 sudo systemctl enable kdump.service
 echo 1 > /proc/sys/kernel/sysrq
 echo "c" > /proc/sysrq-trigger

Expected Results

  1. You can SSH successfully into your FCOS system.
  2. The kernel arguments (as seen in the /proc/cmdline file) are as specified in the Ignition config, or via rpm-ostree after the reboot
  3. After FCOS is rebooted you can successfully see the vmcore and other log files created under /var/crash
  4. The linked documentation makes sense, nothing is obviously missing, and the instructions are clear.