broken pipe error

[vikozo]

Hello
I do the Backup of all virtuell Server and got on a couple of Server the following error

vma_queue_write: write error - Broken pipe

what does it mean and how to solve it?

have a nice day
vinc

 

[t.lamprecht]

If you have enough space left, which I simply guess you have, it may be a problem with your NAS/NFS share, I assume you use one as a backup storage?
Look in the logs of your PVE node and of your storage shoulkd help you to find out the problem, or at least where it lies

Take a look at this post from the mailing list, which had similar problems.
http://pve.proxmox.com/pipermail/pve-user/2014-November/008031.html
http://pve.proxmox.com/pipermail/pve-user/2014-November/008033.html

 

[vikozo]

"...enough space left..." is there a option to tell how many backups should be saved?
would it be possible to create a new Server and add Diskspace wich would not be there?

have a nice day
vinc

 

[t.lamprecht]

"...enough space left..." is there a option to tell how many backups should be saved?

You can limit the maximum number of backups on a storage, but automatically deleting the oldest one is not implemented, AFAIK.

would it be possible to create a new Server and add Diskspace wich would not be there?

Yes, it is possible to create a VM or add a disk to a existing one where the size of the disk is bigger then the remaining space, this is the case as for example qcow2 is thin provisioned, which means it's only as big as needed and grows (but not automatically shrinks) as needed.

 

[vikozo]

You can limit the maximum number of backups on a storage, but automatically deleting the oldest one is not implemented, AFAIK.

Hello
Thanks for your fast feedback, so I will delete them today, will this feature be added soon?
have a nice day
vinc

 

[t.lamprecht]

will this feature be added soon

No, it's not currently in work by anyone here, AFAIK. We naturally accept patches, but a lot of caution is needed for such a feature, nothing worse the falsely deleted Backups in case of a failure, I think, it's surely not as trivial as simply deleting the last backup in case of no remaining space.


This is probably a thing were you're better of writing a shell script which suits your needs exactly and is tailored for your backup strategy, then test it a lot and then some more and use that, IMHO.