update: support upgrading a subset of nodes

It can be useful in a large cluster deployment to split the upgrade and
only upgrade a group of nodes at a time.

Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2014304
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit e5cf9db2b04f55196d867f5a7248b455307f4407)

diff --git a/infrastructure-playbooks/rolling_update.yml b/infrastructure-playbooks/rolling_update.yml
index 30e0862f518994b1bfc554fe62fa9570ea14deb6..1708835555af7fd253073367e1895d5bc1971239 100644 (file)
--- a/infrastructure-playbooks/rolling_update.yml
+++ b/infrastructure-playbooks/rolling_update.yml
@@ -16,6 +16,7 @@
 
 - name: confirm whether user really meant to upgrade the cluster
   hosts: localhost
+  tags: always
   become: false
   gather_facts: false
   vars:
@@ -39,7 +40,6 @@
 
 
 - name: gather facts and check the init system
-
   hosts:
     - "{{ mon_group_name|default('mons') }}"
     - "{{ osd_group_name|default('osds') }}"
@@ -51,7 +51,7 @@
     - "{{ client_group_name|default('clients') }}"
     - "{{ iscsi_gw_group_name|default('iscsigws') }}"
     - "{{ grafana_server_group_name|default('grafana-server') }}"
-
+  tags: always
   any_errors_fatal: True
   become: True
   gather_facts: False
@@ -128,6 +128,7 @@
         rolling_update: true
 
 - name: upgrade ceph mon cluster
+  tags: mons
   vars:
     health_mon_check_retries: 5
     health_mon_check_delay: 15
@@ -307,6 +308,7 @@
 
 - name: reset mon_host
   hosts: "{{ mon_group_name|default('mons') }}"
+  tags: always
   become: True
   gather_facts: false
   tasks:
@@ -324,6 +326,7 @@
     health_mon_check_delay: 15
     upgrade_ceph_packages: True
   hosts: "{{ mon_group_name|default('mons') }}"
+  tags: mgrs
   serial: 1
   become: True
   gather_facts: false
@@ -359,6 +362,7 @@
     upgrade_ceph_packages: True
     ceph_release: "{{ ceph_stable_release }}"
   hosts: "{{ mgr_group_name|default('mgrs') }}"
+  tags: mgrs
   serial: 1
   become: True
   gather_facts: false
@@ -394,6 +398,7 @@
 
 - name: set osd flags
   hosts: "{{ mon_group_name | default('mons') }}[0]"
+  tags: osds
   become: True
   gather_facts: false
   tasks:
@@ -455,8 +460,8 @@
     health_osd_check_retries: 40
     health_osd_check_delay: 30
     upgrade_ceph_packages: True
-
   hosts: "{{ osd_group_name|default('osds') }}"
+  tags: osds
   serial: 1
   become: True
   gather_facts: false
@@ -538,6 +543,7 @@
 
 - name: complete osd upgrade
   hosts: "{{ mon_group_name|default('mons') }}[0]"
+  tags: osds
   become: True
   gather_facts: false
   tasks:
@@ -579,6 +585,7 @@
 
 - name: upgrade ceph mdss cluster, deactivate all rank > 0
   hosts: "{{ mon_group_name | default('mons') }}[0]"
+  tags: mdss
   become: true
   gather_facts: false
   tasks:
@@ -680,6 +687,7 @@
   vars:
     upgrade_ceph_packages: True
   hosts: active_mdss
+  tags: mdss
   become: true
   gather_facts: false
   tasks:
@@ -726,6 +734,7 @@
   vars:
     upgrade_ceph_packages: True
   hosts: standby_mdss
+  tags: mdss
   become: True
   gather_facts: false
 
@@ -774,6 +783,7 @@
   vars:
     upgrade_ceph_packages: True
   hosts: "{{ rgw_group_name|default('rgws') }}"
+  tags: rgws
   serial: 1
   become: True
   gather_facts: false
@@ -818,6 +828,7 @@
   vars:
     upgrade_ceph_packages: True
   hosts: "{{ rbdmirror_group_name|default('rbdmirrors') }}"
+  tags: rbdmirrors
   serial: 1
   become: True
   gather_facts: false
@@ -851,6 +862,7 @@
   vars:
     upgrade_ceph_packages: True
   hosts: "{{ nfs_group_name|default('nfss') }}"
+  tags: nfss
   serial: 1
   become: True
   gather_facts: false
@@ -899,8 +911,8 @@
 - name: upgrade ceph iscsi gateway node
   vars:
     upgrade_ceph_packages: True
-  hosts:
-    - "{{ iscsi_gw_group_name|default('iscsigws') }}"
+  hosts: "{{ iscsi_gw_group_name|default('iscsigws') }}"
+  tags: iscsigws
   serial: 1
   become: True
   gather_facts: false
@@ -942,6 +954,7 @@
   vars:
     upgrade_ceph_packages: True
   hosts: "{{ client_group_name|default('clients') }}"
+  tags: clients
   serial: "{{ client_update_batch | default(20) }}"
   become: True
   gather_facts: false
@@ -973,6 +986,9 @@
     - "{{ rgw_group_name | default('rgws') }}"
     - "{{ rbdmirror_group_name | default('rbdmirrors') }}"
     - "{{ mgr_group_name | default('mgrs') }}"
+  tags:
+    - post_upgrade
+    - crash
   gather_facts: false
   become: true
   tasks:
@@ -1009,6 +1025,7 @@
   - "{{ rbdmirror_group_name | default('rbdmirrors') }}"
   - "{{ client_group_name | default('clients') }}"
   - "{{ iscsi_gw_group_name | default('iscsigws') }}"
+  tags: post_upgrade
   become: True
   gather_facts: false
   tasks:
@@ -1068,6 +1085,7 @@
     - "{{ nfs_group_name|default('nfss') }}"
     - "{{ iscsi_gw_group_name|default('iscsigws') }}"
     - "{{ grafana_server_group_name|default('grafana-server') }}"
+  tags: monitoring
   gather_facts: false
   become: true
   tasks:
@@ -1098,6 +1116,7 @@
 
 - name: upgrade monitoring node
   hosts: "{{ grafana_server_group_name }}"
+  tags: monitoring
   gather_facts: false
   become: true
   tasks:
@@ -1129,6 +1148,7 @@
 
 - name: upgrade ceph dashboard
   hosts: "{{ groups[mgr_group_name] | default(groups[mon_group_name]) | default(omit) }}"
+  tags: monitoring
   gather_facts: false
   become: true
   tasks:
@@ -1148,6 +1168,7 @@
 
 - name: switch any existing crush buckets to straw2
   hosts: "{{ mon_group_name | default('mons') }}[0]"
+  tags: post_upgrade
   become: true
   any_errors_fatal: true
   gather_facts: false
@@ -1189,6 +1210,7 @@
 
 - name: show ceph status
   hosts: "{{ mon_group_name|default('mons') }}"
+  tags: always
   become: True
   gather_facts: false
   tasks: