add max-connect-errors config parameter
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack Percona Cluster Charm |
Fix Released
|
Medium
|
Nicholas Njihia |
Bug Description
Per https:/
* max_connect_errors can have a default value of 10 or 100.
* number of OpenStack API workers can be high if the number of vcpus is, too (and worker-multiplier value is set to the default, 2.0)
We've seen in a couple of environments that MySQL blocks connections from a host after a config change on such host, via Juju. When such re-configuration happens, OpenStack API services restart and the servers try to reconnect to MySQL.
"mysqladmin flush host" is the only action that will resolve the situation.
* Request:
Be able to tweak max_connect_errors value. If change is done when the MySQL service is clustered and running, "set global max_connect_
Changed in charm-percona-cluster: | |
status: | New → Triaged |
importance: | Undecided → Medium |
Changed in charm-percona-cluster: | |
assignee: | nobody → Nicholas Njihia (nicknjihian) |
Changed in charm-percona-cluster: | |
milestone: | none → 21.10 |
Changed in charm-percona-cluster: | |
status: | Fix Committed → Fix Released |
Fix proposed to branch: master /review. opendev. org/c/openstack /charm- percona- cluster/ +/806237
Review: https:/