Page MenuHomePhabricator

Switchover es5 master (es1023 -> es1024)
Closed, ResolvedPublic

Description

When: Anytime, writes will be disabled

Checklist:

NEW primary: es1024
OLD primary: es1023

  • Check configuration differences between new and old primary:
sudo pt-config-diff --defaults-file /root/.my.cnf h=es1024.eqiad.wmnet h=es1023.eqiad.wmnet

Failover prep:

  • Silence alerts on all hosts:
sudo cookbook sre.hosts.downtime --hours 1 -r "Primary switchover es5 T322187" 'A:db-section-es5'
  • Set NEW primary with weight 0 (and depool it from API or vslow/dump groups if it is present).
sudo dbctl instance es1024 set-weight 0
sudo dbctl config commit -m "Set es1024 with weight 0 T361548"
  • Topology changes, move all replicas under NEW primary
sudo db-switchover --timeout=25 --only-slave-move es1023 es1024
  • Disable puppet on both nodes
sudo cumin 'es1023* or es1024*' 'disable-puppet "primary switchover T361548"'

Failover:

  • Log the failover:
!log Starting es5 eqiad failover from es1023 to es1024 T361548
  • Switch primaries:
sudo db-switchover --skip-slave-move es1023 es1024
echo "===== es1023 (OLD)"; sudo db-mysql es1023 -e 'show slave status\G'
echo "===== es1024 (NEW)"; sudo db-mysql es1024 -e 'show slave status\G'
  • Promote NEW primary in dbctl, and remove read-only
sudo dbctl --scope eqiad section es5 set-master es1024
sudo dbctl config commit -m "Promote es1024 to es5 primary T361548"
  • Restart puppet on both hosts:
sudo cumin 'es1023* or es1024*' 'run-puppet-agent -e "primary switchover T361548"'

Clean up tasks:

  • Clean up heartbeat table(s).
sudo db-mysql es1024 heartbeat -e "delete from heartbeat where file like 'es1023%';"
  • change events for query killer:
events_coredb_master.sql on the new primary es1024
events_coredb_slave.sql on the new slave es1023
sudo dbctl instance es1023 set-candidate-master --section es5 true
sudo dbctl instance es1024 set-candidate-master --section es5 false
(dborch1001): sudo orchestrator-client -c untag -i es1024 --tag name=candidate
(dborch1001): sudo orchestrator-client -c tag -i es1023 --tag name=candidate
sudo db-mysql db1115 zarcillo -e "select * from masters where section = 'es5';"
  • (If needed): Depool es1023 for maintenance.
sudo dbctl instance es1023 depool
sudo dbctl config commit -m "Depool es1023 T361548"
  • Change es1023 weight to mimic the previous weight es1024:
sudo dbctl instance es1023 edit
  • Update/resolve this ticket.

Event Timeline

Marostegui triaged this task as Medium priority.Apr 2 2024, 7:40 AM
Marostegui updated the task description. (Show Details)
Marostegui moved this task from Triage to Ready on the DBA board.

Change #1024997 had a related patch set uploaded (by Marostegui; author: Marostegui):

[operations/mediawiki-config@master] db-production.php: Disable writes on es5

https://gerrit.wikimedia.org/r/1024997

Change #1024997 merged by jenkins-bot:

[operations/mediawiki-config@master] db-production.php: Disable writes on es5

https://gerrit.wikimedia.org/r/1024997

Mentioned in SAL (#wikimedia-operations) [2024-04-29T05:08:37Z] <marostegui@deploy1002> Started scap: Backport for [[gerrit:1024997|db-production.php: Disable writes on es5 (T361548)]]

Change #1024998 had a related patch set uploaded (by Marostegui; author: Marostegui):

[operations/puppet@production] mariadb: Promote es1024 to es5 master

https://gerrit.wikimedia.org/r/1024998

Change #1024999 had a related patch set uploaded (by Marostegui; author: Marostegui):

[operations/dns@master] wmnet: Update es5-master CNAME

https://gerrit.wikimedia.org/r/1024999

Mentioned in SAL (#wikimedia-operations) [2024-04-29T05:22:32Z] <marostegui@deploy1002> marostegui: Backport for [[gerrit:1024997|db-production.php: Disable writes on es5 (T361548)]] synced to the testservers (https://wikitech.wikimedia.org/wiki/Mwdebug)

Mentioned in SAL (#wikimedia-operations) [2024-04-29T05:23:11Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Set es1024 with weight 0 T361548', diff saved to https://phabricator.wikimedia.org/P61303 and previous config saved to /var/cache/conftool/dbconfig/20240429-052311-root.json

Mentioned in SAL (#wikimedia-operations) [2024-04-29T05:35:35Z] <marostegui@deploy1002> Finished scap: Backport for [[gerrit:1024997|db-production.php: Disable writes on es5 (T361548)]] (duration: 26m 58s)

Change #1024998 merged by Marostegui:

[operations/puppet@production] mariadb: Promote es1024 to es5 master

https://gerrit.wikimedia.org/r/1024998

Change #1025001 had a related patch set uploaded (by Marostegui; author: Marostegui):

[operations/puppet@production] es1023: Disable notifications

https://gerrit.wikimedia.org/r/1025001

Mentioned in SAL (#wikimedia-operations) [2024-04-29T05:40:08Z] <marostegui> Starting es5 eqiad failover from es1023 to es1024 T361548

Mentioned in SAL (#wikimedia-operations) [2024-04-29T05:40:36Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Promote es1024 to es5 primary T361548', diff saved to https://phabricator.wikimedia.org/P61305 and previous config saved to /var/cache/conftool/dbconfig/20240429-054035-marostegui.json

Change #1024999 merged by Marostegui:

[operations/dns@master] wmnet: Update es5-master CNAME

https://gerrit.wikimedia.org/r/1024999

Mentioned in SAL (#wikimedia-operations) [2024-04-29T05:41:58Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Depool es1023 T361548', diff saved to https://phabricator.wikimedia.org/P61306 and previous config saved to /var/cache/conftool/dbconfig/20240429-054158-marostegui.json

Change #1025001 merged by Marostegui:

[operations/puppet@production] es1023: Disable notifications

https://gerrit.wikimedia.org/r/1025001

Marostegui updated the task description. (Show Details)

All done