Page MenuHomePhabricator

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

Description

When: Anytime, writes will be disabled

Checklist:

NEW primary: es1023
OLD primary: es1024

  • 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 T337285" '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 es1023 set-weight 10
sudo dbctl config commit -m "Set es1023 with weight 0 T337285"
  • Topology changes, move all replicas under NEW primary
sudo db-switchover --timeout=25 --only-slave-move es1024 es1023
  • Disable puppet on both nodes
sudo cumin 'es1023* or es1024*' 'disable-puppet "primary switchover T337285"'

Failover:

  • Log the failover:
!log Starting es5 eqiad failover from es1024 to es1023 T337285
  • Switch primaries:
sudo db-switchover --skip-slave-move es1024 es1023
echo "===== es1024 (OLD)"; sudo db-mysql es1024 -e 'show slave status\G'
echo "===== es1023 (NEW)"; sudo db-mysql es1023 -e 'show slave status\G'
  • Promote NEW primary in dbctl, and remove read-only
sudo dbctl --scope eqiad section es5 set-master es1023
sudo dbctl config commit -m "Promote es1023 to es5 primary T337285"
  • Restart puppet on both hosts:
sudo cumin 'es1023* or es1024*' 'run-puppet-agent -e "primary switchover T337285"'

Clean up tasks:

  • Clean up heartbeat table(s).
sudo db-mysql es1023 heartbeat -e "delete from heartbeat where file like 'es1024%';"
  • change events for query killer:
events_coredb_master.sql on the new primary es1024
events_coredb_slave.sql on the new slave es1024
sudo dbctl instance es1024 set-candidate-master --section es5 true
sudo dbctl instance es1023 set-candidate-master --section es5 false
(dborch1001): sudo orchestrator-client -c untag -i es1023 --tag name=candidate
(dborch1001): sudo orchestrator-client -c tag -i es1024 --tag name=candidate
sudo db-mysql db1215 zarcillo -e "select * from masters where section = 'es5';"
  • (If needed): Depool es1024 for maintenance.
sudo dbctl instance es1024 depool
sudo dbctl config commit -m "Depool es1024 T337285"
  • Change es1024 weight to mimic the previous weight es1023:
sudo dbctl instance es1024 edit
  • Update/resolve this ticket.

Event Timeline

Marostegui updated the task description. (Show Details)

Change 922457 had a related patch set uploaded (by Marostegui; author: Marostegui):

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

https://gerrit.wikimedia.org/r/922457

Change 922458 had a related patch set uploaded (by Marostegui; author: Marostegui):

[operations/puppet@production] mariadb: Promote es1023 to es5 master

https://gerrit.wikimedia.org/r/922458

Change 922459 had a related patch set uploaded (by Marostegui; author: Marostegui):

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

https://gerrit.wikimedia.org/r/922459

Change 922459 merged by jenkins-bot:

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

https://gerrit.wikimedia.org/r/922459

Mentioned in SAL (#wikimedia-operations) [2023-05-23T07:17:49Z] <marostegui@cumin1001> START - Cookbook sre.hosts.downtime for 1:00:00 on 6 hosts with reason: Primary switchover es5 T337285

Mentioned in SAL (#wikimedia-operations) [2023-05-23T07:17:53Z] <marostegui@deploy1002> Started scap: Backport for [[gerrit:922459|db-production.php: Disable writes in es5 (T337285)]]

Mentioned in SAL (#wikimedia-operations) [2023-05-23T07:18:05Z] <marostegui@cumin1001> END (PASS) - Cookbook sre.hosts.downtime (exit_code=0) for 1:00:00 on 6 hosts with reason: Primary switchover es5 T337285

Change 922458 merged by Marostegui:

[operations/puppet@production] mariadb: Promote es1023 to es5 master

https://gerrit.wikimedia.org/r/922458

Mentioned in SAL (#wikimedia-operations) [2023-05-23T07:19:19Z] <marostegui@deploy1002> marostegui: Backport for [[gerrit:922459|db-production.php: Disable writes in es5 (T337285)]] synced to the testservers: mwdebug1002.eqiad.wmnet, mwdebug2001.codfw.wmnet, mwdebug2002.codfw.wmnet, mwdebug1001.eqiad.wmnet

Mentioned in SAL (#wikimedia-operations) [2023-05-23T07:25:09Z] <marostegui@deploy1002> Finished scap: Backport for [[gerrit:922459|db-production.php: Disable writes in es5 (T337285)]] (duration: 07m 16s)

Mentioned in SAL (#wikimedia-operations) [2023-05-23T07:36:17Z] <marostegui> Starting es5 eqiad failover from es1024 to es1023 T337285

Mentioned in SAL (#wikimedia-operations) [2023-05-23T07:37:10Z] <marostegui@cumin1001> dbctl commit (dc=all): 'Promote es1023 to es5 primary T337285', diff saved to https://phabricator.wikimedia.org/P48477 and previous config saved to /var/cache/conftool/dbconfig/20230523-073710-root.json

Change 922457 merged by Marostegui:

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

https://gerrit.wikimedia.org/r/922457

Mentioned in SAL (#wikimedia-operations) [2023-05-23T07:38:41Z] <marostegui@cumin1001> dbctl commit (dc=all): 'Depool es1024 T337285', diff saved to https://phabricator.wikimedia.org/P48479 and previous config saved to /var/cache/conftool/dbconfig/20230523-073841-root.json