Page MenuHomePhabricator

Switchover s1 master (db2103 -> db2112)
Closed, ResolvedPublic

Description

When: During a pre-defined DBA maintenance windows

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

  • Team calendar invite

Affected wikis:: https://noc.wikimedia.org/conf/highlight.php?file=dblists/s1.dblist

Checklist:

NEW primary: db2112
OLD primary: db2103

  • Check configuration differences between new and old primary:
sudo pt-config-diff --defaults-file /root/.my.cnf h=db2103.codfw.wmnet h=db2112.codfw.wmnet

Failover prep:

  • Silence alerts on all hosts:
sudo cookbook sre.hosts.downtime --hours 1 -r "Primary switchover s1 T356059" 'A:db-section-s1'
  • Set NEW primary with weight 0 (and depool it from API or vslow/dump groups if it is present).
sudo dbctl instance db2112 set-weight 0
sudo dbctl config commit -m "Set db2112 with weight 0 T356059"
  • Topology changes, move all replicas under NEW primary
sudo db-switchover --timeout=25 --only-slave-move db2103 db2112
  • Disable puppet on both nodes
sudo cumin 'db2103* or db2112*' 'disable-puppet "primary switchover T356059"'

Failover:

  • Log the failover:
!log Starting s1 codfw failover from db2103 to db2112 - T356059
  • Set section read-only:
sudo dbctl --scope eqiad section s1 ro "Maintenance until 06:15 UTC - T356059"
sudo dbctl --scope codfw section s1 ro "Maintenance until 06:15 UTC - T356059"
sudo dbctl config commit -m "Set s1 codfw as read-only for maintenance - T356059"
  • Check s1 is indeed read-only
  • Switch primaries:
sudo db-switchover --skip-slave-move db2103 db2112
echo "===== db2103 (OLD)"; sudo db-mysql db2103 -e 'show slave status\G'
echo "===== db2112 (NEW)"; sudo db-mysql db2112 -e 'show slave status\G'

[]x Promote NEW primary in dbctl, and remove read-only

sudo dbctl --scope codfw section s1 set-master db2112
sudo dbctl --scope eqiad section s1 rw
sudo dbctl --scope codfw section s1 rw
sudo dbctl config commit -m "Promote db2112 to s1 primary and set section read-write T356059"
  • Restart puppet on both hosts:
sudo cumin 'db2103* or db2112*' 'run-puppet-agent -e "primary switchover T356059"'

Clean up tasks:

  • Clean up heartbeat table(s).
sudo db-mysql db2112 heartbeat -e "delete from heartbeat where file like 'db2103%';"
  • change events for query killer:
events_coredb_master.sql on the new primary db2112
events_coredb_slave.sql on the new slave db2103
sudo dbctl instance db2103 set-candidate-master --section s1 true
sudo dbctl instance db2112 set-candidate-master --section s1 false
(dborch1001): sudo orchestrator-client -c untag -i db2112 --tag name=candidate
(dborch1001): sudo orchestrator-client -c tag -i db2103 --tag name=candidate
sudo db-mysql db1215 zarcillo -e "select * from masters where section = 's1';"
  • (If needed): Depool db2103 for maintenance.
sudo dbctl instance db2103 depool
sudo dbctl config commit -m "Depool db2103 T356059"
  • Change db2103 weight to mimic the previous weight db2112:
sudo dbctl instance db2103 edit
  • Apply outstanding schema changes to db2103 (if any)
  • Update/resolve this ticket.

Event Timeline

Change 993469 had a related patch set uploaded (by Gerrit maintenance bot; author: Gerrit maintenance bot):

[operations/puppet@production] mariadb: Promote db2112 to s1 master

https://gerrit.wikimedia.org/r/993469

Change 993470 had a related patch set uploaded (by Gerrit maintenance bot; author: Gerrit maintenance bot):

[operations/dns@master] wmnet: Update s1-master alias

https://gerrit.wikimedia.org/r/993470

Mentioned in SAL (#wikimedia-operations) [2024-01-30T05:43:44Z] <marostegui@cumin1002> START - Cookbook sre.hosts.downtime for 1:00:00 on 36 hosts with reason: Primary switchover s1 T356059

Mentioned in SAL (#wikimedia-operations) [2024-01-30T05:44:11Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Set db2112 with weight 0 T356059', diff saved to https://phabricator.wikimedia.org/P55848 and previous config saved to /var/cache/conftool/dbconfig/20240130-054410-marostegui.json

Mentioned in SAL (#wikimedia-operations) [2024-01-30T05:44:17Z] <marostegui@cumin1002> END (PASS) - Cookbook sre.hosts.downtime (exit_code=0) for 1:00:00 on 36 hosts with reason: Primary switchover s1 T356059

codfw/sectionLoads/s1 live                                         codfw/sectionLoads/s1 generated
[                                                                  [
    {                                                                  {
        "db2103": 0                                                        "db2103": 0
    },                                                                 },
    {                                                                  {
        "db2112": 300,                                                     "db2112": 0,
        "db2116": 200,                                                     "db2116": 200,
        "db2130": 200,                                                     "db2130": 200,
        "db2145": 200,                                                     "db2145": 200,
        "db2146": 350,                                                     "db2146": 350,
        "db2153": 200,                                                     "db2153": 200,

Change 993469 merged by Marostegui:

[operations/puppet@production] mariadb: Promote db2112 to s1 master

https://gerrit.wikimedia.org/r/993469

Mentioned in SAL (#wikimedia-operations) [2024-01-30T06:12:24Z] <marostegui> Starting s1 codfw failover from db2103 to db2112 - T356059

Mentioned in SAL (#wikimedia-operations) [2024-01-30T06:12:44Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Set s1 codfw as read-only for maintenance - T356059', diff saved to https://phabricator.wikimedia.org/P55851 and previous config saved to /var/cache/conftool/dbconfig/20240130-061243-marostegui.json

Mentioned in SAL (#wikimedia-operations) [2024-01-30T06:13:06Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Promote db2112 to s1 primary and set section read-write T356059', diff saved to https://phabricator.wikimedia.org/P55852 and previous config saved to /var/cache/conftool/dbconfig/20240130-061305-marostegui.json

Change 993470 merged by Marostegui:

[operations/dns@master] wmnet: Update s1-master alias

https://gerrit.wikimedia.org/r/993470

Mentioned in SAL (#wikimedia-operations) [2024-01-30T06:15:29Z] <marostegui@cumin1002> dbctl commit (dc=all): 'Depool db2103 T356059', diff saved to https://phabricator.wikimedia.org/P55854 and previous config saved to /var/cache/conftool/dbconfig/20240130-061529-root.json

Marostegui updated the task description. (Show Details)