Calin Don

There will be some changes in the way the services are handled in the helm chart which will land in v0.6.0 in a few days, that I think will solve...

I think it's related to #615 which is fixed is in the master branch and will be release in 0.5.1.

Please make sure you are not hitting #170. (see https://www.bitpoke.io/docs/mysql-operator/deploy-mysql-cluster/#note-1). Also please try with v0.5.0.

I need to take a look but seems to be a bug. Can you tell how many replicas per cluster?

I don't understand all the implications right now, as I'm not up to date with Pod Security Standards. Can you try a PR and check that it still runs on...

Please try version 0.5.0 and let us know if this still happens. Also please note that we'll drop support for Helm v2 in 0.6.0 (see #731).

The `MySQLCluster` controller/resource should not be responsible for creating any additional services (as is the case of the k8s StatefulSet controller for example). The extra services should be created by...

The `fsGroup` is set in https://github.com/bitpoke/mysql-operator/blob/18a6031621733203b43374b835eaa2fe59d35795/pkg/controller/mysqlcluster/internal/syncer/statefullset.go#L132-L150 If there is a mismatch, means that percona changed user for their containers and we need to account for that in the operator. As...

@SF2311 this still happens with 0.5.0?

It's not recommended to run MySQL on NFS (at least not on v3 which is the version Filestore provides). See https://dev.mysql.com/doc/refman/8.0/en/disk-issues.html#disk-issues-nfs. You should use MySQL Operator with google cloud's persistent...