Page MenuHomePhabricator

Switchover s7 master (db1136 -> db1181)
Closed, ResolvedPublic

Description

When: During a pre-defined DBA maintenance windows - Tuesday 19th April 06:00 AM UTC

  • Team calendar invite

Affected wikis:: https://noc.wikimedia.org/conf/highlight.php?file=dblists/s7.dblist + centralauth

Checklist:

NEW primary: db1181
OLD primary: db1136

  • Check configuration differences between new and old primary:
sudo pt-config-diff --defaults-file /root/.my.cnf h=db1136.eqiad.wmnet h=db1181.eqiad.wmnet

Failover prep:

  • Silence alerts on all hosts:
sudo cookbook sre.hosts.downtime --hours 1 -r "Primary switchover s7 T306001" 'A:db-section-s7'
  • Set NEW primary with weight 0
sudo dbctl instance db1181 set-weight 0
sudo dbctl config commit -m "Set db1181 with weight 0 T306001"
  • Topology changes, move all replicas under NEW primary
sudo db-switchover --timeout=15 --only-slave-move db1136 db1181
  • Disable puppet on both nodes
sudo cumin 'db1136* or db1181*' 'disable-puppet "primary switchover T306001"'

Failover:

  • Log the failover:
!log Starting s7 eqiad failover from db1136 to db1181 - T306001
  • Set section read-only:
sudo dbctl --scope eqiad section s7 ro "Maintenance until 06:15 UTC - T306001"
sudo dbctl config commit -m "Set s7 eqiad as read-only for maintenance - T306001"
  • Check s7 is indeed read-only
  • Switch primaries:
sudo db-switchover --skip-slave-move db1136 db1181
echo "===== db1136 (OLD)"; sudo db-mysql db1136 -e 'show slave status\G'
echo "===== db1181 (NEW)"; sudo db-mysql db1181 -e 'show slave status\G'
  • Promote NEW primary in dbctl, and remove read-only
sudo dbctl --scope eqiad section s7 set-master db1181
sudo dbctl --scope eqiad section s7 rw
sudo dbctl config commit -m "Promote db1181 to s7 primary and set section read-write T306001"
  • Restart puppet on both hosts:
sudo cumin 'db1136* or db1181*' 'run-puppet-agent -e "primary switchover T306001"'

Clean up tasks:

  • Clean up heartbeat table(s).
  • change events for query killer:
events_coredb_master.sql on the new primary db1181
events_coredb_slave.sql on the new slave db1136
sudo dbctl instance db1136 set-candidate-master --section s7 true
sudo dbctl instance db1181 set-candidate-master --section s7 false
(dborch1001): sudo orchestrator-client -c untag -i db1181 --tag name=candidate
(dborch1001): sudo orchestrator-client -c tag -i db1136 --tag name=candidate
sudo dbctl instance db1136 depool
sudo dbctl config commit -m "Depool db1136 T306001"
  • Apply outstanding schema changes to db1136 (if any) T306001#7863050
  • Update/resolve this ticket.

Event Timeline

Marostegui triaged this task as Medium priority.
Marostegui updated Other Assignee, added: Ladsgroup.
Marostegui added projects: DBA, User-notice.
Marostegui updated the task description. (Show Details)
Marostegui updated the task description. (Show Details)

db1181 needs to be rebooted for T303174

Mentioned in SAL (#wikimedia-operations) [2022-04-13T05:35:27Z] <marostegui@cumin1001> dbctl commit (dc=all): 'Depool db1181 for reboot T306001', diff saved to https://phabricator.wikimedia.org/P24557 and previous config saved to /var/cache/conftool/dbconfig/20220413-053526-root.json

db1181 needs to be rebooted for T303174

Done

root@db1181:~# uname -v
#1 SMP Debian 5.10.106-1 (2022-03-17)

Change 783819 had a related patch set uploaded (by Marostegui; author: Marostegui):

[operations/puppet@production] mariadb: Promote db1181 to s7 master

https://gerrit.wikimedia.org/r/783819

Change 783820 had a related patch set uploaded (by Marostegui; author: Marostegui):

[operations/dns@master] wmnet: Update s7-master CNAME

https://gerrit.wikimedia.org/r/783820

Mentioned in SAL (#wikimedia-operations) [2022-04-19T04:57:44Z] <root@cumin1001> START - Cookbook sre.hosts.downtime for 2:00:00 on 25 hosts with reason: Primary switchover s7 T306001

Mentioned in SAL (#wikimedia-operations) [2022-04-19T04:58:01Z] <root@cumin1001> END (PASS) - Cookbook sre.hosts.downtime (exit_code=0) for 2:00:00 on 25 hosts with reason: Primary switchover s7 T306001

Mentioned in SAL (#wikimedia-operations) [2022-04-19T04:58:14Z] <marostegui@cumin1001> dbctl commit (dc=all): 'Set db1181 with weight 0 T306001', diff saved to https://phabricator.wikimedia.org/P25245 and previous config saved to /var/cache/conftool/dbconfig/20220419-045814-root.json

Change 783819 merged by Marostegui:

[operations/puppet@production] mariadb: Promote db1181 to s7 master

https://gerrit.wikimedia.org/r/783819

Mentioned in SAL (#wikimedia-operations) [2022-04-19T06:01:43Z] <marostegui> Starting s7 eqiad failover from db1136 to db1181 - T306001

Mentioned in SAL (#wikimedia-operations) [2022-04-19T06:01:58Z] <marostegui@cumin1001> dbctl commit (dc=all): 'Set s7 eqiad as read-only for maintenance - T306001', diff saved to https://phabricator.wikimedia.org/P25252 and previous config saved to /var/cache/conftool/dbconfig/20220419-060157-marostegui.json

Mentioned in SAL (#wikimedia-operations) [2022-04-19T06:02:26Z] <marostegui@cumin1001> dbctl commit (dc=all): 'Promote db1181 to s7 primary and set section read-write T306001', diff saved to https://phabricator.wikimedia.org/P25253 and previous config saved to /var/cache/conftool/dbconfig/20220419-060226-marostegui.json

Change 783820 merged by Marostegui:

[operations/dns@master] wmnet: Update s7-master CNAME

https://gerrit.wikimedia.org/r/783820

Mentioned in SAL (#wikimedia-operations) [2022-04-19T06:05:59Z] <marostegui@cumin1001> dbctl commit (dc=all): 'Depool db1136 T306001', diff saved to https://phabricator.wikimedia.org/P25254 and previous config saved to /var/cache/conftool/dbconfig/20220419-060559-marostegui.json

Marostegui updated the task description. (Show Details)

Switchover was done, read only times
Start: 06:01:58
Stop: 06:02:26
Total read only time: 28 seconds

The pending schema changes will be tracked on their different tasks, so closing this.

Change 784681 had a related patch set uploaded (by Ladsgroup; author: Amir Sarabadani):

[operations/puppet@production] mariadb: Promote db1104 to s8 master

https://gerrit.wikimedia.org/r/784681