Page MenuHomePhabricator

Switchover s6 master (db2129 -> db2214)
Closed, ResolvedPublic

Description

When: During a pre-defined DBA maintenance windows

Affected wikis:: https://noc.wikimedia.org/conf/highlight.php?file=dblists/s6.dblist

Checklist:

NEW primary: db2214
OLD primary: db2129

  • Check configuration differences between new and old primary:
sudo pt-config-diff --defaults-file /root/.my.cnf h=db2129.codfw.wmnet h=db2214.codfw.wmnet

Failover prep:

  • Silence alerts on all hosts:
sudo cookbook sre.hosts.downtime --hours 1 -r "Primary switchover s6 T374087" 'A:db-section-s6'
  • Set NEW primary with weight 0
sudo dbctl instance db2214 set-weight 0
sudo dbctl config commit -m "Set db2214 with weight 0 T374087"
  • Depool NEW from any specific group (API, vslow, dump) if present.
sudo dbctl instance db2214 edit
# If some changes were made:
sudo dbctl config commit -m "Remove db2214 from API/vslow/dump T374087"
  • Topology changes, move all replicas under NEW primary
sudo db-switchover --timeout=25 --replicating-master --read-only-master --only-slave-move db2129 db2214
  • Disable puppet on both nodes
sudo cumin 'db2129* or db2214*' 'disable-puppet "primary switchover T374087"'
  • Merge gerrit puppet change to promote NEW primary: FIXME

Failover:

  • Log the failover:
!log Starting s6 codfw failover from db2129 to db2214 - T374087
  • Switch primaries:
sudo db-switchover --replicating-master --read-only-master --skip-slave-move db2129 db2214
echo "===== db2129 (OLD)"; sudo db-mysql db2129 -e 'show slave status\G'
echo "===== db2214 (NEW)"; sudo db-mysql db2214 -e 'show slave status\G'
  • Promote NEW primary in dbctl
sudo dbctl --scope codfw section s6 set-master db2214
sudo dbctl config commit -m "Promote db2214 to s6 primary T374087"
  • Restart puppet on both hosts:
sudo cumin 'db2129* or db2214*' 'run-puppet-agent -e "primary switchover T374087"'

Clean up tasks:

  • Clean up heartbeat table(s).
sudo db-mysql db2214 heartbeat -e "delete from heartbeat where file like 'db2129%';"
  • 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 db2214
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 db2129
  • Update candidate primary dbctl and orchestrator notes
sudo dbctl instance db2129 set-candidate-master --section s6 true
sudo dbctl instance db2214 set-candidate-master --section s6 false
sudo cumin 'dborch*' 'orchestrator-client -c untag -i db2214 --tag name=candidate'
sudo cumin 'dborch*' 'orchestrator-client -c tag -i db2129 --tag name=candidate'
sudo db-mysql db1215 zarcillo -e "select * from masters where section = 's6';"
  • (If needed): Depool db2129 for maintenance.
sudo dbctl instance db2129 depool
sudo dbctl config commit -m "Depool db2129 T374087"
  • Change db2129 weight to mimic the previous weight db2214:
sudo dbctl instance db2129 edit
  • Apply outstanding schema changes to db2129 (if any)
  • Update/resolve this ticket.

Details

Event Timeline

Change #1070871 had a related patch set uploaded (by Gerrit maintenance bot; author: Gerrit maintenance bot):

[operations/puppet@production] mariadb: Promote db2214 to s6 master

https://gerrit.wikimedia.org/r/1070871

Ladsgroup moved this task from Triage to In progress on the DBA board.

Mentioned in SAL (#wikimedia-operations) [2024-09-05T11:19:59Z] <ladsgroup@cumin2002> START - Cookbook sre.hosts.downtime for 1:00:00 on 24 hosts with reason: Primary switchover s6 T374087

Mentioned in SAL (#wikimedia-operations) [2024-09-05T11:20:37Z] <ladsgroup@cumin2002> END (PASS) - Cookbook sre.hosts.downtime (exit_code=0) for 1:00:00 on 24 hosts with reason: Primary switchover s6 T374087

Mentioned in SAL (#wikimedia-operations) [2024-09-05T11:21:22Z] <ladsgroup@cumin2002> dbctl commit (dc=all): 'Set db2214 with weight 0 T374087', diff saved to https://phabricator.wikimedia.org/P68688 and previous config saved to /var/cache/conftool/dbconfig/20240905-112121-ladsgroup.json

Replica moves is done in two minutes instead of 25 \o/

Change #1070871 merged by Ladsgroup:

[operations/puppet@production] mariadb: Promote db2214 to s6 master

https://gerrit.wikimedia.org/r/1070871

Mentioned in SAL (#wikimedia-operations) [2024-09-05T11:27:48Z] <Amir1> Starting s6 codfw failover from db2129 to db2214 - T374087

Mentioned in SAL (#wikimedia-operations) [2024-09-05T11:28:48Z] <ladsgroup@cumin2002> dbctl commit (dc=all): 'Promote db2214 to s6 primary T374087', diff saved to https://phabricator.wikimedia.org/P68689 and previous config saved to /var/cache/conftool/dbconfig/20240905-112846-ladsgroup.json

Mentioned in SAL (#wikimedia-operations) [2024-09-05T11:31:55Z] <ladsgroup@cumin2002> dbctl commit (dc=all): 'Depool db2129 T374087', diff saved to https://phabricator.wikimedia.org/P68690 and previous config saved to /var/cache/conftool/dbconfig/20240905-113153-ladsgroup.json

Ladsgroup updated the task description. (Show Details)
Ladsgroup changed the edit policy from "Custom Policy" to "All Users".

Replica moves is done in two minutes instead of 25 \o/

omg