Disks GUI - /proc/mounts too long - aborting

[valeech]

Running pve 5.2-2 (and also previous versions)

When I select the "Disks" option in the GUI I get an error

file '/proc/mounts' too long - aborting (500)​

On the nodes where I receive this message, I have a large number of ZFS datasets with subsequent mountpoints. I believe this is interfering with the Disk display in the GUI. If I remove most of the ZFS datasets, there error goes away.

 

[Stoiko Ivanov]

Seems you hit a limit in our codebase.
Could you open a bugreport on https://bugzilla.proxmox.com/ ?
also could you please paste the output of `wc /proc/mounts` (when the ZFS datasets are mounted) - in the bug-report

Thanks!

 

[kroem]

Running pve 5.2-2 (and also previous versions)

When I select the "Disks" option in the GUI I get an error

file '/proc/mounts' too long - aborting (500)​

On the nodes where I receive this message, I have a large number of ZFS datasets with subsequent mountpoints. I believe this is interfering with the Disk display in the GUI. If I remove most of the ZFS datasets, there error goes away.

I have the same problem, did you file a bug that I can attach to maybe?

 

[mira]

The limit is increased in libpve-common-perl 5.0-41 and newer. See https://bugzilla.proxmox.com/show_bug.cgi?id=1311
If you're on a newer version and still run into this error, please post the output of 'wc /proc/mounts' as Stoiko mentioned above.

 

[kroem]

Actaully this is fine after a reboot it seems, but the output was (before reboot):
root@cat:~# wc /proc/mounts
5966 35796 1072539 /proc/mounts

Right now it is
root@cat:~# wc /proc/mounts
137 822 12225 /proc/mounts

 

[mira]

Could you post the content of /proc/mounts the next time you run into this limit? It would be interesting to see what's filling the file this much.

 

[unixgeek]

I'm also running into this issue:

root@shv0:~# wc /proc/mounts
5679 34074 622502 /proc/mounts

I've tried working around this issue by updating the limit in ProcFSTools.pm based on the git commit in the bugzilla report:

    return PVE::Tools::file_get_contents("/proc/mounts", 768*1024);

But after restarting pveproxy, this doesn't seem to help.
Is there something I can do short of restarting the entire server?

EDIT: I'm using Proxmox 6.0 (libpve-common-perl 6.0-5).

 

[jobboman]

Just experienced this on PVE 8.0.4

wc /proc/mounts:

32812 196872 9786166 /proc/mounts

Both my server node and shared NFS storage are experiencing high load from other tasks running at the same time.

Edit: There was a CIFS mount that was looping and trying to mount over itself forever slowing everything down. I manually unmounted and remounted it and now things appear to be smoothly working.

 

[Stoiko Ivanov]

Edit: There was a CIFS mount that was looping and trying to mount over itself forever slowing everything down. I manually unmounted and remounted it and now things appear to be smoothly working.

If possible please open a fresh thread (and link to this one) in the future - it's better than to reply to one that has seen no activity in almost 4 years...

If you want to investigate this further - please post the journal of the node where this happened to the new thread - and also what tried to mount the cifs-share in a loop