SMB storagebox Hetzner `cloud not activate storage`

[Enthylsa]

Hi, i have a 3 node cluster on root server at Hetzner and using a storage box from Hetzner via SMB /CIFS.

I made an update on December 16 on all three nodes (also some proxmox components have been updated.

Since then I have problems with the storage. First I saw errors from my backup jobs. I tried to access the storage via PVE GUI but it was not accessible and the storage showed a question mark.

i removed it via GUI and tried to add again but this failed with Error 500 (Storage not online).

I saw in syslog of my three nodes that they all still tries to connect to the storage but fails.

Dec 20 12:56:07 server4 kernel: [3467762.218172] CIFS: Status code returned 0xc000006d STATUS_LOGON_FAILURE
Dec 20 12:56:07 server4 kernel: [3467762.218178] CIFS: VFS: \\u320087.your-storagebox.de Send error in SessSetup = -13

I then restarted all tree nodes and this message was gone. Also I was now able to add the storage again and access it via GUI. Everything seems find but still the backup jobs failed with same error message:

TASK ERROR: could not activate storage 'storagebox-proxmox1': storage 'storagebox-proxmox1' does not exist could not get storage information for 'storagebox-proxmox1': storage 'storagebox-proxmox1' does not exist

But, this time I am still able to access the storage via GUI and see the backups iso images etc.

anyone has an idea why this happen and how to solve it?

 

[Ruza]

found a solution? having the same problem

 

[LnxBil]

Try to mount it by hand and enable debug to see a more verbose error message. STATUS_LOGON_FAILURE is often an error in the credentials and/or used smb/cifs version.

 

[Ruza]

Problem is that this error message appears several hours after directory has been successfully mounted and worked with.

 

[LnxBil]

Problem is that this error message appears several hours after directory has been successfully mounted and worked with.

Oh, that's even weirder. Does an immediate remount work or is the share just stale at that moment?

 

[Ruza]

Yes, remount solves the problem for a while. I updated to the latest versions and rebooted the host, no error for 6 hours now

 

[LnxBil]

In the end, you may have to enable debug for cifs (or the mount itself) so that it would log to the kernel ring buffer. Maybe there is something in there that helps.