Page MenuHomePhabricator

Switchover es6 master (es2037 -> es2035)
Closed, ResolvedPublic

Description

When: Anytime - no in use

Prerequisites: https://wikitech.wikimedia.org/wiki/MariaDB/Primary_switchover

Checklist:

NEW primary: es2035
OLD primary: es2037

  • Check configuration differences between new and old primary:
sudo pt-config-diff --defaults-file /root/.my.cnf h=es2037.codfw.wmnet h=es2035.codfw.wmnet

Failover prep:

  • Silence alerts on all hosts:
sudo cookbook sre.hosts.downtime --hours 1 -r "Primary switchover es6 T400436" 'A:db-section-es6'
  • Set NEW primary with weight 0 (and depool it from API or vslow/dump groups if it is present).
sudo dbctl instance es2035 set-weight 0
sudo dbctl config commit -m "Set es2035 with weight 0 T400436"
  • Topology changes, move all replicas under NEW primary
sudo db-switchover --read-only-master --replicating-master --timeout=25 --only-slave-move es2037 es2035
  • Disable puppet on both nodes
sudo cumin 'es2037* or es2035*' 'disable-puppet "primary switchover T400436"'

Failover:

  • Log the failover:
!log Starting es6 codfw failover from es2037 to es2035 - T400436
  • Switch primaries:
sudo db-switchover --read-only-master --replicating-master  --skip-slave-move es2037 es2035
echo "===== es2037 (OLD)"; sudo db-mysql es2037 -e 'show slave status\G'
echo "===== es2035 (NEW)"; sudo db-mysql es2035 -e 'show slave status\G'
  • Promote NEW primary in dbctl, and remove read-only
sudo dbctl --scope codfw section es6 set-master es2035
sudo dbctl config commit -m "Promote es2035 to es6 primary T400436"
  • Clean up heartbeat table(s).
sudo db-mysql es2035 heartbeat -e "delete from heartbeat where file like 'es2037%';"
  • Restart puppet on both hosts:
sudo cumin 'es2037* or es2035*' 'run-puppet-agent -e "primary switchover T400436"'

Clean up tasks:

  • change events for query killer:
curl -sS 'https://gerrit.wikimedia.org/r/plugins/gitiles/operations/software/+/refs/heads/master/dbtools/events_coredb_master.sql?format=TEXT' | base64 -d | sudo db-mysql es2035
curl -sS 'https://gerrit.wikimedia.org/r/plugins/gitiles/operations/software/+/refs/heads/master/dbtools/events_coredb_slave.sql?format=TEXT' | base64 -d | sudo db-mysql es2037
  • Update candidate primary dbctl and orchestrator notes
sudo dbctl instance es2037 set-candidate-master --section es6 true
sudo dbctl instance es2035 set-candidate-master --section es6 false
sudo cumin 'dborch*' 'orchestrator-client -c untag -i es2035 --tag name=candidate'
sudo cumin 'dborch*' 'orchestrator-client -c tag -i es2037 --tag name=candidate'
sudo db-mysql db1215 zarcillo -e "select * from masters where section = 'es6';"
  • (If needed): Depool es2037 for maintenance.
sudo dbctl instance es2037 depool
sudo dbctl config commit -m "Depool es2037 T400436"
  • Change es2037 weight to mimic the previous weight es2035:
sudo dbctl instance es2037 edit
  • Update/resolve this ticket.

Event Timeline

Change #1172512 had a related patch set uploaded (by Gerrit maintenance bot; author: Gerrit maintenance bot):

[operations/puppet@production] mariadb: Promote es2035 to es6 master

https://gerrit.wikimedia.org/r/1172512

Marostegui triaged this task as Medium priority.
Marostegui moved this task from Triage to In progress on the DBA board.

Mentioned in SAL (#wikimedia-operations) [2025-07-25T05:57:44Z] <marostegui@cumin1002> DONE (PASS) - Cookbook sre.hosts.downtime (exit_code=0) for 1:00:00 on 8 hosts with reason: Primary switchover es6 T400436

Mentioned in SAL (#wikimedia-operations) [2025-07-25T05:57:49Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Set es2035 with weight 0 T400436', diff saved to https://phabricator.wikimedia.org/P79891 and previous config saved to /var/cache/conftool/dbconfig/20250725-055749-root.json

Change #1172512 merged by Marostegui:

[operations/puppet@production] mariadb: Promote es2035 to es6 master

https://gerrit.wikimedia.org/r/1172512

Mentioned in SAL (#wikimedia-operations) [2025-07-25T06:00:06Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Promote es2035 to es6 primary T400436', diff saved to https://phabricator.wikimedia.org/P79892 and previous config saved to /var/cache/conftool/dbconfig/20250725-060005-marostegui.json

Mentioned in SAL (#wikimedia-operations) [2025-07-25T06:01:04Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Depool es2037 T400436', diff saved to https://phabricator.wikimedia.org/P79893 and previous config saved to /var/cache/conftool/dbconfig/20250725-060103-root.json

Marostegui updated the task description. (Show Details)

Done

Mentioned in SAL (#wikimedia-operations) [2025-07-25T06:02:46Z] <marostegui> Starting es6 codfw failover from es2037 to es2035 - T400436

Mentioned in SAL (#wikimedia-operations) [2025-07-28T06:30:39Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Depool es2038 T400436', diff saved to https://phabricator.wikimedia.org/P80008 and previous config saved to /var/cache/conftool/dbconfig/20250728-063039-root.json