Hello,
I have been happily using PBS for a while and really enjoy the software.
One pain point is that occasionally my backups fail with either:
which fails the job for an entire node, or:
which fails just that particular VM's backup
In either case, it's somewhat expected for there to be occasional network failures -- it's sending the data over the internet to a remote data store in a different city, and when dealing with large amounts of data over the internet there could reasonably be connection interruptions or faults, etc.
What's surprising to me is proxmox does not appear to retry & resume a failed connection in either of the above failure cases. Normally for web queries, it's common practice to retry any given web query a few times before giving up and considering it a permanent failure.
Proxmox does provide a
I have been happily using PBS for a while and really enjoy the software.
One pain point is that occasionally my backups fail with either:
TASK ERROR: could not activate storage '<name removed>': <name removed>: error fetching datastores - 500 Can't connect to <ip removed:port>which fails the job for an entire node, or:
Code:
INFO: 17% (8.5 GiB of 50.0 GiB) in 4m 44s, read: 98.8 MiB/s, write: 95.3 MiB/s
INFO: 18% (9.0 GiB of 50.0 GiB) in 4m 49s, read: 105.8 MiB/s, write: 101.1 MiB/s
INFO: 18% (9.3 GiB of 50.0 GiB) in 10m 19s, read: 721.8 KiB/s, write: 697.5 KiB/s
ERROR: job failed with err -110 - Connection timed out
INFO: aborting backup job
INFO: resuming VM again
INFO: removing (old) fleecing image 'local-zfs:vm-1016-fleece-0'
ERROR: Backup of VM 1016 failed - job failed with err -110 - Connection timed out
INFO: Failed at 2026-07-18 05:16:36which fails just that particular VM's backup
In either case, it's somewhat expected for there to be occasional network failures -- it's sending the data over the internet to a remote data store in a different city, and when dealing with large amounts of data over the internet there could reasonably be connection interruptions or faults, etc.
What's surprising to me is proxmox does not appear to retry & resume a failed connection in either of the above failure cases. Normally for web queries, it's common practice to retry any given web query a few times before giving up and considering it a permanent failure.
Proxmox does provide a
Repeat missed option, but this is for missed as opposed to connection failure events and so I don't think it applies here(?). Could proxmox add a retry option in future?
Last edited:
