Change the default for rabbitmq back to ha-mode: all

In change Ib62001c03e1e08f58cf0c6e0ba07a8879a584084 we switched the
rabbitmq queues HA mode from ha-all to ha-exactly. While this gives us a
nice performance boost with rabbitmq, it makes rabbit less resilient to
network glitches as we painfully found out via
https://bugzilla.redhat.com/show_bug.cgi?id=1441635.

This is the THT part of the change that changes the default to
ha-mode: all.

Closes-Bug: #1686337
Co-Authored-By: Damien Ciabrini <dciabrin@redhat.com>
Co-Authored-By: John Eckersberg <jeckersb@redhat.com>
Change-Id: I7afcf2b3c8deb13fc2134e4cae9c06a44e775384
Depends-On: I9a90e71094b8d8d58b5be0a45a2979701b0ac21c

diff --git a/puppet/services/pacemaker/rabbitmq.yaml b/puppet/services/pacemaker/rabbitmq.yaml
index caada95..30ea6d6 100644 (file)
--- a/puppet/services/pacemaker/rabbitmq.yaml
+++ b/puppet/services/pacemaker/rabbitmq.yaml
@@ -39,34 +39,5 @@ outputs:
           - rabbitmq::service_manage: false
       step_config: |
         include ::tripleo::profile::pacemaker::rabbitmq
-      upgrade_tasks:
-        - name: get bootstrap nodeid
-          tags: common
-          command: hiera bootstrap_nodeid
-          register: bootstrap_node
-        - name: set is_bootstrap_node fact
-          tags: common
-          set_fact: is_bootstrap_node={{bootstrap_node.stdout == ansible_hostname}}
-        - name: get rabbitmq policy
-          tags: common
-          shell: pcs resource show rabbitmq | grep -q -E "Attributes:.*\"ha-mode\":\"all\""
-          register: rabbit_ha_mode
-          when: is_bootstrap_node
-          ignore_errors: true
-        - name: set migrate_rabbit_ha_mode fact
-          tags: common
-          set_fact: migrate_rabbit_ha_mode={{rabbit_ha_mode.rc == 0}}
-          when: is_bootstrap_node
-        - name: Fixup for rabbitmq ha-queues LP#1668600
-          tags: step0,pre-upgrade
-          shell: |
-            nr_controllers=$(($(hiera controller_node_names | grep -o "," |wc -l) + 1))
-            nr_queues=$(($nr_controllers / 2 + ($nr_controllers % 2)))
-            if ! [ $nr_queues -gt 0 -a $nr_queues -le $nr_controllers ]; then
-                echo "ERROR: The nr. of HA queues during the rabbit upgrade is out of range: $nr_queues"
-                exit 1
-            fi
-            pcs resource update rabbitmq set_policy='ha-all ^(?!amq\\.).* {"ha-mode":"exactly","ha-params":'"$nr_queues}" --wait=600
-          when: is_bootstrap_node and migrate_rabbit_ha_mode
       metadata_settings:
         get_attr: [RabbitMQServiceBase, role_data, metadata_settings]

diff --git a/puppet/services/rabbitmq.yaml b/puppet/services/rabbitmq.yaml
index 4747978..d69da3e 100644 (file)
--- a/puppet/services/rabbitmq.yaml
+++ b/puppet/services/rabbitmq.yaml
@@ -40,10 +40,10 @@ parameters:
     hidden: true
   RabbitHAQueues:
     description:
-      The number of HA queues to be configured in rabbit. The default is 0 which will
-      be automatically overridden to CEIL(N/2) where N is the number of nodes running
-      rabbitmq.
-    default: 0
+      The number of HA queues to be configured in rabbit. The default is -1 which
+      translates to "ha-mode all". The special value 0 will be automatically
+      overridden to CEIL(N/2) where N is the number of nodes running rabbitmq.
+    default: -1
     type: number
   MonitoringSubscriptionRabbitmq:
     default: 'overcloud-rabbitmq'

diff --git a/releasenotes/notes/change-rabbitmq-ha-mode-policy-default-6c6cd7f02181f0e0.yaml b/releasenotes/notes/change-rabbitmq-ha-mode-policy-default-6c6cd7f02181f0e0.yaml
new file mode 100644 (file)
index 0000000..d6f74ef
--- /dev/null
+++ b/releasenotes/notes/change-rabbitmq-ha-mode-policy-default-6c6cd7f02181f0e0.yaml
@@ -0,0 +1,11 @@
+---
+upgrade:
+  - |
+    We are not changing the rabbitmq ha-mode policy during upgrades any longer.
+    The policy chosen at deploy time will remain the same but can be changed
+    manually.
+fixes:
+  - |
+    Due to https://bugs.launchpad.net/tripleo/+bug/1686337 we switch the
+    default of rabbitmq back ha-mode "all". This is to make the installation
+    more robust in the face of network issues.