When: 11th Nov at 06:00 AM UTC
Checklist:
- Create a task to communicate the chosen date and send an announcement to the community: T294322
- Create a calendar entry for the maintenance, invite sre-data-persistence@
- Add to deployments calendar. E.g.:
{{Deployment calendar event card
|when=2021-11-10 22:00 SF
|length=0.5
|window=Database primary switchover for s8
|who={{ircnick|kormat|Stevie Beth Mhaol}}, {{ircnick|marostegui|Manuel 'Early Bird' Arostegui}}
|what=https://phabricator.wikimedia.org/T294321
}}NEW primary: db1109
OLD primary: db1104
- Check configuration differences between new and old primary:
sudo pt-config-diff --defaults-file /root/.my.cnf h=db1104.eqiad.wmnet h=db1109.eqiad.wmnet
Failover prep:
- Silence alerts on all hosts:
sudo cookbook sre.hosts.downtime --hours 1 -r "Primary switchover s8 T294321" 'A:db-section-s8'
- Set NEW primary with weight 0
sudo dbctl instance db1109 set-weight 0 sudo dbctl config commit -m "Set db1109 with weight 0 T294321"
- Topology changes, move all replicas under NEW primary
sudo db-switchover --timeout=15 --only-slave-move db1104 db1109
- Disable puppet on both nodes
sudo cumin 'db1104* or db1109*' 'disable-puppet "primary switchover T294321"'
- Merge gerrit puppet change to promote NEW primary: https://gerrit.wikimedia.org/r/c/operations/puppet/+/737831
Failover:
- Log the failover:
!log Starting s8 eqiad failover from db1104 to db1109 - T294321
- Set section read-only:
sudo dbctl --scope eqiad section s8 ro "Maintenance until 06:15 UTC - T294321" sudo dbctl config commit -m "Set s8 eqiad as read-only for maintenance - T294321"
- Check s8 is indeed read-only
- Switch primaries:
sudo db-switchover --skip-slave-move db1104 db1109 echo "===== db1104 (OLD)"; sudo mysql.py -h db1104 -e 'show slave status\G' echo "===== db1109 (NEW)"; sudo mysql.py -h db1109 -e 'show slave status\G'
- Promote NEW primary in dbctl, and remove read-only
sudo dbctl --scope eqiad section s8 set-master db1109 sudo dbctl --scope eqiad section s8 rw sudo dbctl config commit -m "Promote db1109 to s8 primary and set section read-write T294321"
- Restart puppet on both hosts:
sudo cumin 'db1104* or db1109*' 'run-puppet-agent -e "primary switchover T294321"'
Clean up tasks:
- Clean up heartbeat table(s).
[x change events for query killer:
events_coredb_master.sql on the new primary db1109 events_coredb_slave.sql on the new slave db1104
- Update DNS: https://gerrit.wikimedia.org/r/c/operations/dns/+/737832
- Update candidate primary dbctl notes
sudo dbctl instance db1104 set-candidate-master --section s8 true sudo dbctl instance db1109 set-candidate-master --section s8 false
- Check tendril was updated
- Check zarcillo was updated
- db-switchover should do this. If it fails, do it manually: https://phabricator.wikimedia.org/P13956
- Depool OLD primary, as it's running 10.1, replicating from a 10.4 primary
sudo dbctl instance db1104 depool sudo dbctl config commit -m "Depool db1104 until it's reimaged to buster T294321"
- Update/resolve this ticket.