When: During a pre-defined DBA maintenance windows
Prerequisites: https://wikitech.wikimedia.org/wiki/MariaDB/Primary_switchover
- Team calendar invite
Affected wikis:: https://noc.wikimedia.org/conf/highlight.php?file=dblists/s7.dblist
Checklist:
NEW primary: db1236
OLD primary: db1181
- Check configuration differences between new and old primary:
sudo pt-config-diff --defaults-file /root/.my.cnf h=db1181.eqiad.wmnet h=db1236.eqiad.wmnet
Failover prep:
- Silence alerts on all hosts:
sudo cookbook sre.hosts.downtime --hours 1 -r "Primary switchover s7 T370121" 'A:db-section-s7'
- Set NEW primary with weight 0
sudo dbctl instance db1236 set-weight 0 sudo dbctl config commit -m "Set db1236 with weight 0 T370121"
- Depool NEW from any specific group (API, vslow, dump) if present.
sudo dbctl instance db1236 edit # If some changes were made: sudo dbctl config commit -m "Remove db1236 from API/vslow/dump T370121"
- Topology changes, move all replicas under NEW primary
sudo db-switchover --timeout=25 --only-slave-move db1181 db1236
- Disable puppet on both nodes
sudo cumin 'db1181* or db1236*' 'disable-puppet "primary switchover T370121"'
- Merge gerrit puppet change to promote NEW primary: https://gerrit.wikimedia.org/r/c/operations/puppet/+/1054413
Failover:
- Log the failover:
!log Starting s7 eqiad failover from db1181 to db1236 - T370121
- Set section read-only:
sudo dbctl --scope eqiad section s7 ro "Maintenance until 06:15 UTC - T370121" sudo dbctl --scope codfw section s7 ro "Maintenance until 06:15 UTC - T370121" sudo dbctl config commit -m "Set s7 eqiad as read-only for maintenance - T370121"
- Check s7 is indeed read-only
- Switch primaries:
sudo db-switchover --skip-slave-move db1181 db1236 echo "===== db1181 (OLD)"; sudo db-mysql db1181 -e 'show slave status\G' echo "===== db1236 (NEW)"; sudo db-mysql db1236 -e 'show slave status\G'
- Promote NEW primary in dbctl, and remove read-only
sudo dbctl --scope eqiad section s7 set-master db1236 sudo dbctl --scope eqiad section s7 rw sudo dbctl --scope codfw section s7 rw sudo dbctl config commit -m "Promote db1236 to s7 primary and set section read-write T370121"
- Restart puppet on both hosts:
sudo cumin 'db1181* or db1236*' 'run-puppet-agent -e "primary switchover T370121"'
Clean up tasks:
- Clean up heartbeat table(s).
sudo db-mysql db1236 heartbeat -e "delete from heartbeat where file like 'db1181%';"
- 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 db1236 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 db1181
- Update DNS: https://gerrit.wikimedia.org/r/c/operations/dns/+/1054414
- Update candidate primary dbctl and orchestrator notes
sudo dbctl instance db1181 set-candidate-master --section s7 true sudo dbctl instance db1236 set-candidate-master --section s7 false sudo cumin 'dborch*' 'orchestrator-client -c untag -i db1236 --tag name=candidate' sudo cumin 'dborch*' 'orchestrator-client -c tag -i db1181 --tag name=candidate'
- Check zarcillo was updated
- db-switchover should do this. If it fails, do it manually: https://phabricator.wikimedia.org/P13956
sudo db-mysql db1215 zarcillo -e "select * from masters where section = 's7';"
- (If needed): Depool db1181 for maintenance.
sudo dbctl instance db1181 depool sudo dbctl config commit -m "Depool db1181 T370121"
- Change db1181 weight to mimic the previous weight db1236:
sudo dbctl instance db1181 edit
- Update/resolve this ticket.