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/s4.dblist
Checklist:
NEW primary: db2179
OLD primary: db2140
- Check configuration differences between new and old primary:
sudo pt-config-diff --defaults-file /root/.my.cnf h=db2140.codfw.wmnet h=db2179.codfw.wmnet
Failover prep:
- Silence alerts on all hosts:
sudo cookbook sre.hosts.downtime --hours 1 -r "Primary switchover s4 T355658" 'A:db-section-s4'
- Set NEW primary with weight 0 (and depool it from API or vslow/dump groups if it is present).
sudo dbctl instance db2179 set-weight 0 sudo dbctl config commit -m "Set db2179 with weight 0 T355658"
- Topology changes, move all replicas under NEW primary
sudo db-switchover --timeout=25 --only-slave-move db2140 db2179
- Disable puppet on both nodes
sudo cumin 'db2140* or db2179*' 'disable-puppet "primary switchover T355658"'
- Merge gerrit puppet change to promote NEW primary: FIXME
Failover:
- Log the failover:
!log Starting s4 codfw failover from db2140 to db2179 - T355658
- Set section read-only:
sudo dbctl --scope eqiad section s4 ro "Maintenance until 06:15 UTC - T355658" sudo dbctl --scope codfw section s4 ro "Maintenance until 06:15 UTC - T355658" sudo dbctl config commit -m "Set s4 codfw as read-only for maintenance - T355658"
- Check s4 is indeed read-only
- Switch primaries:
sudo db-switchover --skip-slave-move db2140 db2179 echo "===== db2140 (OLD)"; sudo db-mysql db2140 -e 'show slave status\G' echo "===== db2179 (NEW)"; sudo db-mysql db2179 -e 'show slave status\G'
- Promote NEW primary in dbctl, and remove read-only
sudo dbctl --scope codfw section s4 set-master db2179 sudo dbctl --scope eqiad section s4 rw sudo dbctl --scope codfw section s4 rw sudo dbctl config commit -m "Promote db2179 to s4 primary and set section read-write T355658"
- Restart puppet on both hosts:
sudo cumin 'db2140* or db2179*' 'run-puppet-agent -e "primary switchover T355658"'
Clean up tasks:
- Clean up heartbeat table(s).
sudo db-mysql db2179 heartbeat -e "delete from heartbeat where file like 'db2140%';"
- change events for query killer:
events_coredb_master.sql on the new primary db2179 events_coredb_slave.sql on the new slave db2140
- Update DNS: FIXME
- Update candidate primary dbctl and orchestrator notes
sudo dbctl instance db2140 set-candidate-master --section s4 true sudo dbctl instance db2179 set-candidate-master --section s4 false (dborch1001): sudo orchestrator-client -c untag -i db2179 --tag name=candidate (dborch1001): sudo orchestrator-client -c tag -i db2140 --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 = 's4';"
- (If needed): Depool db2140 for maintenance.
sudo dbctl instance db2140 depool sudo dbctl config commit -m "Depool db2140 T355658"
- Change db2140 weight to mimic the previous weight db2179:
sudo dbctl instance db2140 edit
- Apply outstanding schema changes to db2140 (if any)
- Update/resolve this ticket.