communication failure (0) / Connection timed out (596)

[felon]

Hello everyone,

i've a running proxmox server and rented a pbs access from a service provider for backups.
Everything was absolutely fine with it.
But now after some time and a lot of more backups, i've the following problem.

i get communication failure (0)" (somtimes a "Connection timed out (596) really seldom) when i try to "list" the backups via gui.
If i list the repository via cli, its working and takes about 40 seconds to list all backups.

Is there any solution for the gui timeout or am i doing something wrong?

Best regards
Ralph

 

[Chris]

Hello everyone,

i've a running proxmox server and rented a pbs access from a service provider for backups.
Everything was absolutely fine with it.
But now after some time and a lot of more backups, i've the following problem.

i get communication failure (0)" (somtimes a "Connection timed out (596) really seldom) when i try to "list" the backups via gui.
If i list the repository via cli, its working and takes about 40 seconds to list all backups.

Is there any solution for the gui timeout or am i doing something wrong?

Best regards
Ralph

Hi, how many backups are we talking about here? Maybe you can reorganize the backups into namespaces so that not all of them have to be fetched/listed from a single api call? See https://pbs.proxmox.com/docs/storage.html#backup-namespaces

 

[felon]

Hi Chris,

thank you very much for the fast reply
after checking i recognized that we have really a lot of backups in the same name-space
thought we set it up with different namespaces but i was wrong.

is it possible to move existing backups into different namespaces? (sorry if its another dumb question)

Best regards
Ralph

 

[Chris]

You could do this by creating the namespace and then moving the backups via a local since job.

 

[felon]

Solution was: working with namespaces as suggested
Thank you for your help

 

[RavenRAMIREZ]

Hello @Chris .
I encounter the same problem. Don't you think it would be wise to change this problem in the development of PBS since a lot of people encounter this problem.

 

[Chris]

Hello @Chris .
I encounter the same problem. Don't you think it would be wise to change this problem in the development of PBS since a lot of people encounter this problem.

Yes, I think we should introduce pagination support for the listing. Please open a bug report at https://bugzilla.proxmox.com/ so we can keep track of the issue and you get notified on status updates, thanks.

 

[Winet.maier]

We are having this issue also here and it makes me some kind of nervous. We split in to namespaces but it will not realy work. So you guys realy have to take care on it. In our case it is the backup from a single vm and I wnat to sea the backups of it and I still get a timeout here.

 

[jcassel]

I am also seeing this same issue. Any progress on getting this fixed?

 

[omgs]

I'm having this issue in a bullseye installation (the pbs is 3.3). It was working and stopped working getting "broken pipe (596)". I have just a few cts in root ns, and the rest in different namespaces.

The pbs backups are unaccesible from pve, but pbs works nice with host backups.

I've deleted the storage and try to recreate it again, but I get:

create storage failed: cfs-lock 'file-storage_cfg' error: got lock request timeout (500)

I guess that it was working well with pbs 3.2-9, and started like this about one week ago, not sure if when upgraded to pbs 3.3.

Also, I'm trying to use the proxmox-backup-client for a pool named 'backup', but I'm unable to make it work because wrong backup specification.

 

[Chris]

For reference: https://bugzilla.proxmox.com/show_bug.cgi?id=3752 and https://bugzilla.proxmox.com/show_bug.cgi?id=3045.

I am also seeing this same issue. Any progress on getting this fixed?

What storage is used by the datastore? Please note that for the listing, the snaphshot contents have to be fetched from the storage, which might take more than the 30s timeout for the api request.

 

[Chris]

create storage failed: cfs-lock 'file-storage_cfg' error: got lock request timeout (500)

This however seems totally unrelated to the initial issue reported in this thread. Check that your PVE node has quorum.

Also, I'm trying to use the proxmox-backup-client for a pool named 'backup', but I'm unable to make it work because wrong backup specification.

Pools are PVE specific, the Proxmox Backup Server is agnostic to these (although one could create a namespace for each pool).

You will have to use vzdump --pool <poolname> --storage <pbs-storage> to backup pools via the cli. See man vzdump for details on available parameters.

 

[omgs]

This however seems totally unrelated to the initial issue reported in this thread. Check that your PVE node has quorum.


Pools are PVE specific, the Proxmox Backup Server is agnostic to these (although one could create a namespace for each pool).

You will have to use vzdump --pool <poolname> --storage <pbs-storage> to backup pools via the cli. See man vzdump for details on available parameters.

Thanks. I've used vzdump from the command line and works perfectly, so it's not a pbs issue. I'll post in pve forum.