Page MenuHomePhabricator

Switchover s6 master (db1201 -> db1173)
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: db1173
OLD primary: db1201

  • Check configuration differences between new and old primary:
sudo pt-config-diff --defaults-file /root/.my.cnf h=db1201.eqiad.wmnet h=db1173.eqiad.wmnet

Failover prep:

  • Silence alerts on all hosts:
sudo cookbook sre.hosts.downtime --hours 1 -r "Primary switchover s6 T387433" 'A:db-section-s6'
  • Set NEW primary with weight 0
sudo dbctl instance db1173 set-weight 0
sudo dbctl config commit -m "Set db1173 with weight 0 T387433"
  • Depool NEW from any specific group (API, vslow, dump) if present.
sudo dbctl instance db1173 edit
# If some changes were made:
sudo dbctl config commit -m "Remove db1173 from API/vslow/dump T387433"
  • Topology changes, move all replicas under NEW primary
sudo db-switchover --timeout=25 --replicating-master --read-only-master --only-slave-move db1201 db1173
  • Disable puppet on both nodes
sudo cumin 'db1201* or db1173*' 'disable-puppet "primary switchover T387433"'

Failover:

  • Log the failover:
!log Starting s6 eqiad failover from db1201 to db1173 - T387433
  • Switch primaries:
sudo db-switchover --replicating-master --read-only-master --skip-slave-move db1201 db1173
echo "===== db1201 (OLD)"; sudo db-mysql db1201 -e 'show slave status\G'
echo "===== db1173 (NEW)"; sudo db-mysql db1173 -e 'show slave status\G'
  • Promote NEW primary in dbctl
sudo dbctl --scope eqiad section s6 set-master db1173
sudo dbctl config commit -m "Promote db1173 to s6 primary T387433"
  • Clean up heartbeat table(s).
sudo db-mysql db1173 heartbeat -e "delete from heartbeat where file like 'db1201%';"
  • Restart puppet on both hosts:
sudo cumin 'db1201* or db1173*' 'run-puppet-agent -e "primary switchover T387433"'

Clean up tasks:

  • 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 db1173
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 db1201
  • Update candidate primary dbctl and orchestrator notes
sudo dbctl instance db1201 set-candidate-master --section s6 true
sudo dbctl instance db1173 set-candidate-master --section s6 false
sudo cumin 'dborch*' 'orchestrator-client -c untag -i db1173 --tag name=candidate'
sudo cumin 'dborch*' 'orchestrator-client -c tag -i db1201 --tag name=candidate'
sudo db-mysql db1215 zarcillo -e "select * from masters where section = 's6';"
  • (If needed): Depool db1201 for maintenance.
sudo dbctl instance db1201 depool
sudo dbctl config commit -m "Depool db1201 T387433"
  • Change db1201 weight to mimic the previous weight db1173:
sudo dbctl instance db1201 edit
  • Update/resolve this ticket.

Details

Event Timeline

Change #1123276 had a related patch set uploaded (by Gerrit maintenance bot; author: Gerrit maintenance bot):

[operations/puppet@production] mariadb: Promote db1173 to s6 master

https://gerrit.wikimedia.org/r/1123276

Marostegui triaged this task as Medium priority.
Marostegui moved this task from Triage to Ready on the DBA board.

@Ladsgroup this will happen tomorrow or Monday, in case you have schema changes pending on the future old eqiad s6 master.

Marostegui added a parent task: Restricted Task.

Mentioned in SAL (#wikimedia-operations) [2025-02-27T13:02:41Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Set db1173 with weight 0 T387433', diff saved to https://phabricator.wikimedia.org/P73765 and previous config saved to /var/cache/conftool/dbconfig/20250227-130240-marostegui.json

Mentioned in SAL (#wikimedia-operations) [2025-02-27T13:02:45Z] <marostegui@cumin1002> DONE (PASS) - Cookbook sre.hosts.downtime (exit_code=0) for 1:00:00 on 24 hosts with reason: Primary switchover s6 T387433

Mentioned in SAL (#wikimedia-operations) [2025-02-27T13:03:14Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Remove db1173 from API/vslow/dump T387433', diff saved to https://phabricator.wikimedia.org/P73766 and previous config saved to /var/cache/conftool/dbconfig/20250227-130313-marostegui.json

Change #1123276 merged by Marostegui:

[operations/puppet@production] mariadb: Promote db1173 to s6 master

https://gerrit.wikimedia.org/r/1123276

Mentioned in SAL (#wikimedia-operations) [2025-02-27T13:11:59Z] <marostegui> Starting s6 eqiad failover from db1201 to db1173 - T387433

Mentioned in SAL (#wikimedia-operations) [2025-02-27T13:12:19Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Promote db1173 to s6 primary T387433', diff saved to https://phabricator.wikimedia.org/P73770 and previous config saved to /var/cache/conftool/dbconfig/20250227-131218-marostegui.json

Mentioned in SAL (#wikimedia-operations) [2025-02-27T13:13:10Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Depool db1201 T387433', diff saved to https://phabricator.wikimedia.org/P73771 and previous config saved to /var/cache/conftool/dbconfig/20250227-131310-marostegui.json