enable kdump

Signed-off-by: Poorna Gottimukkula <[email protected]>

Author: Poorna-Gottimukkula1

examples/install_vars.yaml

@@ -38,3 +38,12 @@ chronyconfig:
        options: iburst
 
 # bastion_vip: "B.B.B.B"
+# For enabling the kdump 
+kdump:
+  enabled: false
+  commandline_remove: "hugepages hugepagesz slub_debug quiet log_buf_len swiotlb"
+  commandline_append: "irqpoll maxcpus=1 reset_devices cgroup_disable=memory mce=off numa=off udev.children-max=2 panic=10 rootflags=nofail acpi_no_memhotplug transparent_hugepage=never nokaslr novmcoredd hest_disable srcutree.big_cpu_lim=0"
+  kexec_args: "-s"
+  kdump_img: "vmlinuz"
+  log_path: "/var/crash"
+  crash_kernel_memory: "2G-4G:384M,4G-16G:512M,16G-64G:1G,64G-128G:2G,128G-:4G"

playbooks/roles/ocp-config/README.md

@@ -56,6 +56,13 @@ Role Variables
 | luks.options                | no       | ["--cipher", "aes-cbc-essiv:sha256"]  | Set List of luks options for the luks encryption |
 | luks.wipeVolume             | no       | true                                  | Configures the luks encrypted partition to be wiped |
 | luks.name                   | no       | root                                  | Set the value for the user label of Filesystem to be luks encrypted |
+| kdump.enabled               | no       | false                                 | Set to true to enable the kdump on Cluster Nodes in OCP deployment |
+| kdump.commandline_remove    | no       | "hugepages hugepagesz slub_debug quiet log_buf_len swiotlb" | This option removes arguments from the current kdump command line |
+| kdump.commandline_append    | no       | "irqpoll maxcpus=1 reset_devices cgroup_disable=memory mce=off numa=off udev.children-max=2 panic=10 rootflags=nofail acpi_no_memhotplug transparent_hugepage=never nokaslr novmcoredd hest_disable srcutree.big_cpu_lim=0" | This option appends arguments to the current kdump command line |
+| kdump.kexec_args            | no       | "-s"                                  | For adding any extra argument to pass to kexec command |
+| kdump.kdump_img             | no       | "vmlinuz"                             | For specifying image other than default kernel image |
+| kdump.log_path              | no       | "/var/crash"                          | The file system path in which the kdump saves the vmcore file |
+| kdump.crash_kernel_memory   | no       | "2G-4G:384M,4G-16G:512M,16G-64G:1G,64G-128G:2G,128G-:4G" | The memory reservation for kdump occurs during the system boot |
 
 *chronyconfig variable example *
 
@@ -90,6 +97,17 @@ luks:
  wipeVolume: true
  name: root
 ```
+*Ansible variable for enabling the Kdump example*
+```
+kdump:
+  enabled: true
+  commandline_remove: "hugepages hugepagesz slub_debug quiet log_buf_len swiotlb"
+  commandline_append: "irqpoll maxcpus=1 reset_devices cgroup_disable=memory mce=off numa=off udev.children-max=2 panic=10 rootflags=nofail acpi_no_memhotplug transparent_hugepage=never nokaslr novmcoredd hest_disable srcutree.big_cpu_lim=0"
+  kexec_args: "-s"
+  kdump_img: "vmlinuz"
+  log_path: "/var/crash"
+  crash_kernel_memory: "2G-4G:384M,4G-16G:512M,16G-64G:1G,64G-128G:2G,128G-:4G"
+```
 
 Dependencies
 ------------

playbooks/roles/ocp-config/defaults/main/main.yaml

@@ -30,3 +30,11 @@ luks:
       - aes-cbc-essiv:sha256
    wipeVolume: true
    name: root
+kdump:
+  enabled: false
+  commandline_remove: "hugepages hugepagesz slub_debug quiet log_buf_len swiotlb"
+  commandline_append: "irqpoll maxcpus=1 reset_devices cgroup_disable=memory mce=off numa=off udev.children-max=2 panic=10 rootflags=nofail acpi_no_memhotplug transparent_hugepage=never nokaslr novmcoredd hest_disable srcutree.big_cpu_lim=0"
+  kexec_args: "-s"
+  kdump_img: "vmlinuz"
+  log_path: "/var/crash"
+  crash_kernel_memory: "2G-4G:384M,4G-16G:512M,16G-64G:1G,64G-128G:2G,128G-:4G"

playbooks/roles/ocp-config/tasks/main.yaml

@@ -99,6 +99,26 @@
       - "/machineconfig/*-chrony-configuration.yaml"
     when: chronyconfig.enabled
 
+  - name: Generate kdump machine configuration
+    vars:
+      kdump_sysconfig: |
+        KDUMP_COMMANDLINE_REMOVE="{{ kdump.commandline_remove }}"
+        KDUMP_COMMANDLINE_APPEND="{{ kdump.commandline_append }}"
+        KEXEC_ARGS="{{ kdump.kexec_args }}"
+        KDUMP_IMG="{{ kdump.kdump_img }}"
+      kdump_conf: |
+        path {{ kdump.log_path }}
+        core_collector makedumpfile -l --message-level 7 -d 31    
+      crash_kernel_memory:  "{{ kdump.crash_kernel_memory }}"
+      role_n: "{{ item }}"
+    template:
+      src: ../templates/99-kdump-node-machineconfig.yaml.j2
+      dest: "{{ workdir }}/manifests/99-kdump-{{ item }}-node-machineconfig.yaml"
+    with_items:
+      - worker
+      - master
+    when: kdump.enabled
+
   - name: Create ignition files
     shell: "OPENSHIFT_INSTALL_RELEASE_IMAGE_OVERRIDE={{ release_image_override }} openshift-install create ignition-configs --log-level {{ log_level }}"
     args:

playbooks/roles/ocp-config/templates/99-kdump-node-machineconfig.yaml.j2

@@ -0,0 +1,28 @@
+apiVersion: machineconfiguration.openshift.io/v1
+kind: MachineConfig
+metadata:
+  labels:
+    machineconfiguration.openshift.io/role: '{{ role_n }}'
+  name: 99-{{ role_n }}-kdump
+spec:
+  config:
+    ignition:
+      version: 3.2.0
+    storage:
+      files:
+        - contents:
+            source: data:text/plain;charset=utf-8;base64,{{ kdump_conf|b64encode }}
+          mode: 420
+          overwrite: true
+          path: /etc/kdump.conf
+        - contents:
+            source: data:text/plain;charset=utf-8;base64,{{ kdump_sysconfig|b64encode }}
+          mode: 420
+          overwrite: true
+          path: /etc/sysconfig/kdump
+    systemd:
+      units:
+        - enabled: true
+          name: kdump.service
+  kernelArguments:
+    - crashkernel="{{ crash_kernel_memory }}"