Sheng Yang

Maybe we can utilize `dm-integrity` for this.

@shangma will continue to work on this one. We need it for v1.2.0 otherwise our support work after will be pretty difficult.

Move to v1.2.1 since the code change is too big for v1.2.0 after RC.

Update on this: https://github.com/rancher/support-bundle-kit now features `simulator` command to mock a Kubernetes cluster, making it extremely easy to look into the k8s environment for troubleshooting.

@Sheldonwl I think the user can already do this in v1.1.0, as a part of https://github.com/longhorn/longhorn/issues/1180 , e.g. https://longhorn.io/docs/1.1.0/monitoring/integrating-with-rancher-monitoring/ Can you confirm it's done?

@amnk Unfortunately I am not super familiar with this part. @PhanLe1010 what do you think?

@IJOL You can setup a recurring snapshot/backup for each volume which will automatically clean up system generated snapshot as a side effect. See https://longhorn.io/docs/1.0.2/best-practices/#volume-maintenance

Bad cron config? Can you elaborate?

Regarding the automatically cleanup system generated snapshot, we're planning to include a feature in v1.1.0 to help. #1669 Regarding saving the space for replica rebuilding, we're working #1304 which is...

@courcelm Can you elaborate your case of accumulating snapshot? What does it look like? Is it system generated (for rebuild) or not?