Page MenuHomePhabricator

Switchover es7 master (es1039 -> es1035)
Closed, ResolvedPublic

Description

When: Anytime - no in use

Prerequisites: https://wikitech.wikimedia.org/wiki/MariaDB/Primary_switchover

Checklist:

NEW primary: es1035
OLD primary: es1039

  • Check configuration differences between new and old primary:
sudo pt-config-diff --defaults-file /root/.my.cnf h=es1039.eqiad.wmnet h=es1035.eqiad.wmnet

Failover prep:

  • Silence alerts on all hosts:
sudo cookbook sre.hosts.downtime --hours 1 -r "Primary switchover es7 T387271" 'A:db-section-es7'
  • Set NEW primary with weight 0 (and depool it from API or vslow/dump groups if it is present).
sudo dbctl instance es1035 set-weight 0
sudo dbctl config commit -m "Set es1035 with weight 0 T387271"
  • Topology changes, move all replicas under NEW primary
sudo db-switchover --read-only-master --replicating-master --timeout=25 --only-slave-move es1039 es1035
  • Disable puppet on both nodes
sudo cumin 'es1039* or es1035*' 'disable-puppet "primary switchover T387271"'

Failover:

  • Log the failover:
!log Starting es7 eqiad failover from es1039 to es1035 - T387271
  • Switch primaries:
sudo db-switchover --read-only-master --replicating-master  --skip-slave-move es1039 es1035
echo "===== es1039 (OLD)"; sudo db-mysql es1039 -e 'show slave status\G'
echo "===== es1035 (NEW)"; sudo db-mysql es1035 -e 'show slave status\G'
  • Promote NEW primary in dbctl, and remove read-only
sudo dbctl --scope eqiad section es7 set-master es1035
sudo dbctl --scope eqiad section es7 rw
sudo dbctl config commit -m "Promote es1035 to es7 primary and set section read-write T387271"
  • Clean up heartbeat table(s).
sudo db-mysql es1035 heartbeat -e "delete from heartbeat where file like 'es1039%';"
  • Restart puppet on both hosts:
sudo cumin 'es1039* or es1035*' 'run-puppet-agent -e "primary switchover T387271"'

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 es1035
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 es1039
  • Update candidate primary dbctl and orchestrator notes
sudo dbctl instance es1039 set-candidate-master --section es7 true
sudo dbctl instance es1035 set-candidate-master --section es7 false
sudo cumin 'dborch*' 'orchestrator-client -c untag -i es1035 --tag name=candidate'
sudo cumin 'dborch*' 'orchestrator-client -c tag -i es1039 --tag name=candidate'
sudo db-mysql db1215 zarcillo -e "select * from masters where section = 'es7';"
  • (If needed): Depool es1039 for maintenance.
sudo dbctl instance es1039 depool
sudo dbctl config commit -m "Depool es1039 T387271"
  • Change es1039 weight to mimic the previous weight es1035:
sudo dbctl instance es1039 edit
  • Update/resolve this ticket.

Details

Event Timeline

Change #1122860 had a related patch set uploaded (by Gerrit maintenance bot; author: Gerrit maintenance bot):

[operations/puppet@production] mariadb: Promote es1035 to es7 master

https://gerrit.wikimedia.org/r/1122860

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

Mentioned in SAL (#wikimedia-operations) [2025-02-26T07:26:00Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Set es1035 with weight 0 T387271', diff saved to https://phabricator.wikimedia.org/P73614 and previous config saved to /var/cache/conftool/dbconfig/20250226-072600-root.json

Mentioned in SAL (#wikimedia-operations) [2025-02-26T07:26:05Z] <marostegui@cumin1002> DONE (PASS) - Cookbook sre.hosts.downtime (exit_code=0) for 1:00:00 on 6 hosts with reason: Primary switchover es7 T387271

Change #1122860 merged by Marostegui:

[operations/puppet@production] mariadb: Promote es1035 to es7 master

https://gerrit.wikimedia.org/r/1122860

Mentioned in SAL (#wikimedia-operations) [2025-02-26T07:27:26Z] <marostegui> Starting es7 eqiad failover from es1039 to es1035 - T387271

Mentioned in SAL (#wikimedia-operations) [2025-02-26T07:27:51Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Promote es1035 to es7 primary and set section read-write T387271', diff saved to https://phabricator.wikimedia.org/P73615 and previous config saved to /var/cache/conftool/dbconfig/20250226-072751-root.json

Mentioned in SAL (#wikimedia-operations) [2025-02-26T07:28:46Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Depool es1039 T387271', diff saved to https://phabricator.wikimedia.org/P73617 and previous config saved to /var/cache/conftool/dbconfig/20250226-072845-root.json