Unexpected pod restart during resize storage size of tikv

Question

Unexpected pod restart during resize storage size of tikv

ideascf opened this issue 6 months ago · 0 comments

Bug Report

Pods of tikv have been restarted during resizing storage size of TiKV component. The impacted cluster is running with tc.spec.configUpdateStrategy: InPlace, and that value used to be tc.spec.configUpdateStrategy: RollingUpdate when the cluster is created. It seems that pvc_modifier tries to recreate tikv STS to do resize storage size. And then tikv_member_manager create a new STS but with a wrong ConfigMap having the same data and another name 'db-tikv' (without hash suffix). The wrongly used ConfigMap causes STS controller to rolling restart all TiKV pods and without leader eviction.

What version of Kubernetes are you using?

v1.28.6-eks-508b6b3

What version of TiDB Operator are you using?

v1.6.0-alpha.11

What storage classes exist in the Kubernetes cluster and what are used for PD/TiKV pods?

It doesn't matter

What's the status of the TiDB cluster pods?

Unexpectedly restart

What did you do?

What did you expect to see?
No restart

What did you see instead?
Restarting