KOS Team

After investigation, we found this bug is an known issue https://mariadb.com/kb/en/transaction-coordinator-log-overview/#you-must-enable-exactly-n-storage-engines. The documentation suggests the following solution:: > To recover from this error, delete the file defined by the [--log-tc](https://mariadb.com/kb/en/mysqld-options/#-log-tc)...

@w3z315 We were able to manually resolve this issue following the steps described in the manual > After investigation, we found this bug is an known issue https://mariadb.com/kb/en/transaction-coordinator-log-overview/#you-must-enable-exactly-n-storage-engines. The documentation...

Hi @mmontes11 thanks for the reply. It seems that the MariaDB pods are able to run in the root mode successfully. It is only the recovery job pods which are...

Thanks for the confirmation. We can take a look into the potential fix

Got it, thanks for the explanation! Perhaps making the property immutable would avoid the confusion?

I see. Should the default value for the `config` get changed to empty string?

@burmanm Thanks for the explanation. I double checked and there are no seed-node labels on the unready Pods. Only one Pod came up and that's the only Pod with the...

@HoustonPutman Thanks for the confirmation. Looking forward!

@csuzhangxc The root cause is because the operator switched to use a client with non-TLS, while the PD is still running with TLS. This causes the calls to the PD...

The ConfigMap is not updated. The pd_member_manager early returns at this line: https://github.com/pingcap/tidb-operator/blob/ff467a6e7a0563b31a3ace2fe5d060774012780d/pkg/manager/member/pd_member_manager.go#L200 which is before where the ConfigMap is updated: https://github.com/pingcap/tidb-operator/blob/ff467a6e7a0563b31a3ace2fe5d060774012780d/pkg/manager/member/pd_member_manager.go#L209 And as mentioned in the issue report, the...