Add introspection rule to update deploy kernel location

Back in the Ussuri release, we changed the Bifrost kernel location to
use ipa.kernel instead of ipa.vmlinuz. While this works fine for newly
discovered nodes, any node added to Bifrost in Train or earlier will
have kept ipa.vmlinuz as its deploy kernel.

This can cause issues since upgrading Bifrost or building new deployment
images will update ipa.kernel, but not ipa.vmlinuz, resulting in nodes
booting with an old kernel and a new ramdisk.

Fix by adding a new rule updating the legacy kernel location and
documenting how to update node information.

Change-Id: I6ae44dcae78424b4638762d015cf24336444707f
(cherry picked from commit 44af704)

ansible/group_vars/all/bifrost

@@ -99,7 +99,7 @@ kolla_bifrost_inspector_port_addition: "{{ inspector_add_ports }}"
 kolla_bifrost_inspector_extra_kernel_options: "{{ inspector_extra_kernel_options }}"
 
 # List of introspection rules for Bifrost's Ironic Inspector service.
-kolla_bifrost_inspector_rules: "{{ inspector_rules }}"
+kolla_bifrost_inspector_rules: "{{ inspector_rules + [inspector_rule_legacy_deploy_kernel] }}"
 
 # Ironic inspector IPMI username to set.
 kolla_bifrost_inspector_ipmi_username: "{{ ipmi_username }}"
@@ -117,6 +117,9 @@ kolla_bifrost_inspector_deploy_kernel: "http://{{ provision_oc_net_name | net_ip
 # Ironic inspector deployment ramdisk location.
 kolla_bifrost_inspector_deploy_ramdisk: "http://{{ provision_oc_net_name | net_ip }}:8080/ipa.initramfs"
 
+# Ironic inspector legacy deployment kernel location.
+kolla_bifrost_inspector_legacy_deploy_kernel: "http://{{ provision_oc_net_name | net_ip }}:8080/ipa.vmlinuz"
+
 # Timeout of hardware inspection on overcloud nodes, in seconds. Default is
 # {{ inspector_inspection_timeout }}.
 kolla_bifrost_inspection_timeout: "{{ inspector_inspection_timeout }}"

ansible/group_vars/all/inspector

@@ -119,6 +119,18 @@ inspector_rule_deploy_kernel:
       path: "driver_info/deploy_kernel"
       value: "{{ inspector_rule_var_deploy_kernel }}"
 
+# Ironic inspector rule to update deployment kernel from legacy location.
+inspector_rule_legacy_deploy_kernel:
+  description: "Update deploy kernel from legacy"
+  conditions:
+    - field: "node://driver_info.deploy_kernel"
+      op: "eq"
+      value: "{{ inspector_rule_var_legacy_deploy_kernel }}"
+  actions:
+    - action: "set-attribute"
+      path: "driver_info/deploy_kernel"
+      value: "{{ inspector_rule_var_deploy_kernel }}"
+
 # Deployment ramdisk referenced by inspector rule.
 inspector_rule_var_deploy_ramdisk:
 

ansible/seed-introspection-rules.yml

@@ -20,3 +20,4 @@
       inspector_rule_var_lldp_switch_port_interface: "{{ kolla_bifrost_inspector_lldp_switch_port_interface }}"
       inspector_rule_var_deploy_kernel: "{{ kolla_bifrost_inspector_deploy_kernel }}"
       inspector_rule_var_deploy_ramdisk: "{{ kolla_bifrost_inspector_deploy_ramdisk }}"
+      inspector_rule_var_legacy_deploy_kernel: "{{ kolla_bifrost_inspector_legacy_deploy_kernel }}"

etc/kayobe/bifrost.yml

@@ -116,6 +116,9 @@
 # Ironic inspector deployment ramdisk location.
 #kolla_bifrost_inspector_deploy_ramdisk:
 
+# Ironic inspector legacy deployment kernel location.
+#kolla_bifrost_inspector_legacy_deploy_kernel:
+
 # Timeout of hardware inspection on overcloud nodes, in seconds. Default is
 # {{ inspector_inspection_timeout }}.
 #kolla_bifrost_inspection_timeout:

releasenotes/notes/update-deploy-kernel-6943f3ad3cb82c51.yaml

@@ -0,0 +1,20 @@
+---
+upgrade:
+  - |
+    Adds an introspection rule to update the location of the deployment kernel
+    registered in existing Ironic nodes. Nodes discovered on a deployment
+    running the Train release or earlier may still be using the ``ipa.vmlinuz``
+    kernel, which stays unchanged when deployment images get updated. If only
+    default introspection rules are in use, existing nodes may be updated from
+    the Bifrost container with the following command:
+
+    ``OS_CLOUD=bifrost baremetal introspection reprocess $NODE_UUID_OR_NAME``
+
+    If non-default rules are used, reprocessing may revert any customisation
+    done by the operator. In this case, a more cautious approach is to update
+    the deployment kernel location manually:
+
+    ``OS_CLOUD=bifrost baremetal node set --driver-info deploy_kernel=<http://url/to/ipa.kernel> $NODE_UUID_OR_NAME``
+
+    If the ``kolla_bifrost_inspector_rules`` list is customised, the rule
+    ``inspector_rule_legacy_deploy_kernel`` should be added to it.