Container Stuck in Rollback and Cannot Start After Unlocking

[ana_mera]

Hey Proxmox Community,

I'm running into an issue where sometimes my containers get stuck during a rollback. When this happens, I stop the rollback manually, but then the container becomes locked. After unlocking it, I'm unable to start the container anymore and have to restart the node to get things working again.

Does anyone know why this might be happening and if there's any other solution besides restarting the node? Any insights or advice would be greatly appreciated!

Thanks!

 

[dakralex]

Could you elaborate on any warnings/errors/logs when the rollback gets stuck and when starting the container fails?

 

[ana_mera]

It doesnt really show any "errors" just gets stuck in rollback and then I stop it manually and also when i try to start it again its stuck and i have to stop it again and then just shows TASK ERROR: zfs error: received interrupt

 

[waltar]

What exactly do you rollback and how, the container, data from outside or inside, an application inside the container ?

 

[ana_mera]

Sorry I wasnt more clear. I rollback a snapshot of my CT

 

[waltar]

I would not expect that rollback of a running container will function, stop it, switch the file and start it.

 

[ana_mera]

Ok, I understand what you mean, next time I will shut it down before I do a rollback.

But now since i stopped the rollback and it wont start is there any way of starting the CT other then restarting the node?

 

[waltar]

A pve node restart should not be required at all, perhaps there was some lost process which hold your container file and so blocked new container start. Take a look at "pct list" for container ("qm list" for vm's) or directly to processes "ps -ef". Perhaps kill with "kill -9 <pid>".