Databases on m3: phabricator
When: Tuesday 15th at 08:00 AM UTC
Impact: Writes will be disabled for around 1 minute.
Failover process
OLD MASTER: db1107
NEW MASTER: db1183
- Check configuration differences between new and old master
$ pt-config-diff h=db1107.eqiad.wmnet,F=/root/.my.cnf h=db1183.eqiad.wmnet,F=/root/.my.cnf
- Silence alerts on all hosts
- Topology changes: move everything under db1183
db-switchover --timeout=15 --only-slave-move db1107.eqiad.wmnet db1183.eqiad.wmnet
- Disable puppet db1107 and db1183 puppet agent --disable "switchover to db1107 T301219"
- Merge gerrit: https://gerrit.wikimedia.org/r/c/operations/puppet/+/762146
- Run puppet on dbproxy1016 and dbproxy1020 and check the config
puppet agent -tv && cat /etc/haproxy/conf.d/db-master.cfg
- Start the failover: !log Failover m3 from db1107 to db1183 - T301219
- Set phabricator in RO:
ssh phab1001
sudo /srv/phab/phabricator/bin/config set cluster.read-only true
# restart database server
sudo /srv/phab/phabricator/bin/config set cluster.read-only false- DB switchover
root@cumin1001:~/wmfmariadbpy/wmfmariadbpy# db-switchover --read-only-master --skip-slave-move db1107 db1183
- Reload haproxies (dbproxy1020 is the active one)
dbproxy1016: systemctl reload haproxy && echo "show stat" | socat /run/haproxy/haproxy.sock stdio dbproxy1020: systemctl reload haproxy && echo "show stat" | socat /run/haproxy/haproxy.sock stdio
- kill connections on the old master (db1107)
pt-kill --print --kill --victims all --match-all F=/dev/null,S=/run/mysqld/mysqld.sock
- Restart puppet on old and new masters (for heartbeat):db1107 and db1183 puppet agent --enable && puppet agent -tv
- Check services affected: phabricator
- Clean orchestrator heartbeat to remove the old masters' one, otherwise Orchestrator will show lag: T301219#7710051
- Close this ticket and create a ticket to update m5: T301654