pvetest
, if you want to try it out.Hello and thanks for the reply . Then the storage space of the storage is sufficient, the percentage occupied is 50%. The transfer of the backups takes place on a dedicated nic, even if in the same net there is an nfs server, which generates little traffic. It only happened once. I am attaching some screenshots.Hi, lamprcht
first a few questions:
* did it work earlier and just happened now/once
* is there higher latency between that Proxmox VE setup and the PBS server?
* What storage is on the PBS (slow spinners or the fast stuff?)
In general, that can indicate an issue when QEMU tries to connect to the PBS in sync.
This whole deal got a lot better in QEMU 5.2, which allows to do some QEMU monitor commands (like this one) asynchrony, avoiding blocking too much.
It is currently available onpvetest
, if you want to try it out.
https://pve.proxmox.com/wiki/Package_Repositories
Hi, unfortunately the event has come backHi,
first a few questions:
* did it work earlier and just happened now/once
* is there higher latency between that Proxmox VE setup and the PBS server?
* What storage is on the PBS (slow spinners or the fast stuff?)
In general, that can indicate an issue when QEMU tries to connect to the PBS in sync.
This whole deal got a lot better in QEMU 5.2, which allows to do some QEMU monitor commands (like this one) asynchrony, avoiding blocking too much.
It is currently available onpvetest
, if you want to try it out.
https://pve.proxmox.com/wiki/Package_Repositories
Can you post the full text log of the problematic VMs or at least parts of that?Hi, unfortunately the event has come back
PBS :Can you post the full text log of the problematic VMs or at least parts of that?
And, I assume this was still using QEMU 5.1?