Page MenuHomePhabricator
Paste P17043

Authored by Kormat on Aug 18 2021, 10:20 AM.
Moved to software/dbtools/

Event Timeline

Kormat edited the content of this paste. (Show Details)

$ ~/ s7 codfw db2118 db2121 T289129


  • Create a task to communicate the chosen date and send an announcement to the community: FIXME
  • Create a calendar entry for the maintenance, invite sre-data-persistence@
  • Add to deployments calendar. E.g.:
{{Deployment calendar event card
    |when=2021-08-24 23:00 SF
    |window=Database primary switchover for s7
    |who={{ircnick|kormat|Stevie Beth Mhaol}}, {{ircnick|marostegui|Manuel 'Early Bird' Arostegui}}

NEW primary: db2121
OLD primary: db2118

  • Check configuration differences between new and old primary:
sudo pt-config-diff --defaults-file /root/.my.cnf h=db2118.codfw.wmnet h=db2121.codfw.wmnet

Failover prep:

  • Silence alerts on all hosts:
sudo cookbook sre.hosts.downtime --hours 1 -r "Primary switchover s7 T289129" 'A:db-section-s7'
  • Set NEW primary with weight 0
sudo dbctl instance db2121 set-weight 0
sudo dbctl config commit -m "Set db2121 with weight 0 T289129"
  • Topology changes, move all replicas under NEW primary
sudo db-switchover --timeout=15 --only-slave-move db2118 db2121
  • Disable puppet on both nodes
sudo cumin 'db2118* or db2121*' 'disable-puppet "primary switchover T289129"'
  • Merge gerrit puppet change to promote NEW primary: FIXME


  • Log the failover:
!log Starting s7 codfw failover from db2118 to db2121 - T289129
  • Set section read-only:
sudo dbctl --scope codfw section s7 ro "Maintenance until 05:15 UTC - T289129"
sudo dbctl config commit -m "Set s7 codfw as read-only for maintenance - T289129"
  • Check s7 is indeed read-only
  • Switch primaries:
sudo db-switchover --skip-slave-move db2118 db2121
echo "===== db2118 (OLD)"; sudo -h db2118 -e 'show slave status\G'
echo "===== db2121 (NEW)"; sudo -h db2121 -e 'show slave status\G'
  • Promote NEW primary in dbctl, and remove read-only
sudo dbctl --scope codfw section s7 set-master db2121
sudo dbctl --scope codfw section s7 rw
sudo dbctl config commit -m "Promote db2121 to s7 primary and set section read-write T289129"
  • Restart puppet on both hosts:
sudo cumin 'db2118* or db2121*' 'run-puppet-agent -e "primary switchover T289129"'

Clean up tasks:

  • Clean up heartbeat table(s).
  • change events for query killer:
events_coredb_master.sql on the new primary db2121
events_coredb_slave.sql on the new slave db2118
  • Update DNS: FIXME
  • Update candidate primary dbctl notes
sudo dbctl instance db2118 set-candidate-master --section s7 true
sudo dbctl instance db2121 set-candidate-master --section s7 false
  • Check tendril was updated
  • Check zarcillo was updated
  • Depool OLD primary, as it's running 10.1, replicating from a 10.4 primary
sudo dbctl instance db2118 depool
sudo dbctl config commit -m "Depool db2118 until it's reimaged to buster T289129"
  • Apply outstanding schema changes to db2118 (if any)
  • Update/resolve this ticket.
Kormat edited the content of this paste. (Show Details)