kubeplus
kubeplus copied to clipboard
Failure encoutered when upgrading KubePlus chart
When upgrading KubePlus chart, encountered the following error:
Warning FailedAttachVolume 79s attachdetach-controller Multi-Attach error for volume "pvc-81a9746b-5bbb-46f2-80f6-70de3f7589f2" Volume is already used by pod(s) kubeplus-deployment-68bc784565-gvrj2
This error shows up in the 'kubectl describe' of the new Pod. The current Pod is not deleted. It is still running.
Per ChatGPT, we can fix this by selecting one of the following strategies:
-
Add terminationGracePeriodSeconds to Deployment: spec: terminationGracePeriodSeconds: 30
-
If you're using a Deployment, fine-tune the rolling update strategy to avoid multiple pods up at once: strategy: type: RollingUpdate rollingUpdate: maxSurge: 0 maxUnavailable: 1