Shuo Wu

The corresponding Longhorn backend API is `/v1/backingimages/{name}/download` And the Longhorn UI PR is: https://github.com/longhorn/longhorn-ui/pull/519 (This PR also includes one upload-API-related modification, which fits the recent backing image refactoring)

> Will it be a problem for an HA distributed system when that single instance manager is not available? I think this should be fine. Assuming the IMs on a...

> * Before the proxy, when the instance manager on the node goes down, backup volume controller still can [read backup](https://github.com/longhorn/longhorn-manager/blob/79ce2cfde922a7e0966059f72b8aee5f05bff513/controller/backup_volume_controller.go#L342) and [delete backup](https://github.com/longhorn/longhorn-manager/blob/79ce2cfde922a7e0966059f72b8aee5f05bff513/controller/backup_controller.go#L255) > * after the proxy, this...

YES. If the instance manager pods on some nodes are unavailable due to some issues like CPU resource exhaustion or wrong tolerations, the backup CRs on the corresponding nodes cannot...

> @shuo-wu did you mean to keep this part "and it will skip the auto-deletion for the failed backups?" Right. This part is what I suggest in the previous comment....

> Ah, sorry. I was not aware of that is your suggestion. Just thought the simple error message is enough for debugging. But yes, we can keep it ("and it...

I may prefer to make this check a separate function, or a part of the replica `Open` function. But due to the complexity of the replica state, more investigation is...

The case instance manager **CR** not found is totally different from the case instance manager **pod** is not found. The error reported above is [here](https://github.com/longhorn/longhorn-manager/blob/8af120dd7c109d81d86118e482f88a26ddf7650f/controller/instance_handler.go#L215) (instance manager CR not found...

I am thinking if we still need to add extra logic to handle [this error ](https://github.com/longhorn/longhorn-manager/blob/8af120dd7c109d81d86118e482f88a26ddf7650f/controller/instance_handler.go#L215). Right now I cannot find another way to trigger this error except for instance...

Found the error in #1975. Let's see if Khushboo can provide the reproducing step first.