Storga Stale file handle, no errore reported

[Valerio Pachera]

Problem:
I scheduled a backup and set to always send an email.
The backup fail to start because of and issue on the backup storage.
I get no mail at all.
Nothing is reported in the tasks table.

More info:
I get an email from cron daemon with subject
Cron <root@pve> vzdump 101 --mailnotification always --mailto '...hidden...' --quiet 1 --node pve --storage cubo --compress lzo --mode snapshot
The mail body contains:
unable to activate storage 'cubo' - directory '/mnt/pve/cubo' does not exist

I checked the mount point and that's what I get:

ls -la /mnt/pve/
ls: cannot access /mnt/pve/cubo: Stale file handle
total 8
drwxr-xr-x 3 root root 4096 Aug 25 16:40 .
drwxr-xr-x 4 root root 4096 Aug 29 12:15 ..
d????????? ? ? ? ? ? cubo

This storage is an nfs storage and the destination server has been rebooted in the past.
I tried to disable and re-enable the storage hoping that proxmox was able to mount but it fails.

I'm using proxmox 4.2-2/725d76f0.

1) My main concern is that proxmox didn't handle the error and reported no message.

To fix the issue I simply rebooted the server.
If you have any suggestion in case it will happen again, please answer back.

2) do you think that I will always get 'Stale file handle' when a nfs storage reboot or goes off-line for some time?

 

[fireon]

Stale file say's there is something corrupt on your storage... or there are wrong rights on. For sending email you must configure your postfix localy with an smarthostentry. The importent with is the sender. an *.local or an *.lan adress is not possible ot send mails to the real internet. If you send only in your lan, this is also working. That is not necessary if you use an public network ipv4 or ipv6 with real domain.

Can you mount your nfsshare from an other host wirteable?

 

[liptech]

Problem:
I scheduled a backup and set to always send an email.
The backup fail to start because of and issue on the backup storage.
I get no mail at all.
Nothing is reported in the tasks table.

More info:
I get an email from cron daemon with subject
Cron <root@pve> vzdump 101 --mailnotification always --mailto '...hidden...' --quiet 1 --node pve --storage cubo --compress lzo --mode snapshot
The mail body contains:
unable to activate storage 'cubo' - directory '/mnt/pve/cubo' does not exist

I checked the mount point and that's what I get:

ls -la /mnt/pve/
ls: cannot access /mnt/pve/cubo: Stale file handle
total 8
drwxr-xr-x 3 root root 4096 Aug 25 16:40 .
drwxr-xr-x 4 root root 4096 Aug 29 12:15 ..
d????????? ? ? ? ? ? cubo

This storage is an nfs storage and the destination server has been rebooted in the past.
I tried to disable and re-enable the storage hoping that proxmox was able to mount but it fails.

I'm using proxmox 4.2-2/725d76f0.

1) My main concern is that proxmox didn't handle the error and reported no message.

To fix the issue I simply rebooted the server.
If you have any suggestion in case it will happen again, please answer back.

2) do you think that I will always get 'Stale file handle' when a nfs storage reboot or goes off-line for some time?

Dear Valerius,
I had the same problem and accidentally solved it by executing the command "umount / mnt / pve / BKP" with BKP being my mount point.
Hope this helps.