Hi there,
we recently added a new server into our cluster of Proxmox VE, this is the first server using AMD cpu. It's a server with 1x AMD Epyc 7302p, 256 gb ram, 2x 1,92 Tb NVMe.
We've a dedicated server with Proxmox Backup Server versino 2.0-14 installed.
On all the other server, using Intel Xeon cpu, we are doing backup to the PBS since a lot of time right now (more than an year) and everything is fine.
Starting using this new AMD Epyc server, we've noticed that is impossible to take a successful backup through PBS. We can just do a local backup or maybe over a NFS storage, but not over PBS datastore.
When we try making a backup over PBS, it seems to start correctly but then it hangs in this way:
And stays in that way for a long time (more than 30 mins) and then it hangs in timeout... server are connected with each other with 10 GbE network and everything works well. I can even restore a backup from PBS to the new server and it works perfectly!
I cannot just do backup over PBS.
In also, with Linux VM, when it hangs in this way, then VM filesystem's got dirty and I must reboot and do fsck /dev/sda1 manually during boot!
No one knows how to fix this?
Thanks in advance...
Paolo
we recently added a new server into our cluster of Proxmox VE, this is the first server using AMD cpu. It's a server with 1x AMD Epyc 7302p, 256 gb ram, 2x 1,92 Tb NVMe.
We've a dedicated server with Proxmox Backup Server versino 2.0-14 installed.
On all the other server, using Intel Xeon cpu, we are doing backup to the PBS since a lot of time right now (more than an year) and everything is fine.
Starting using this new AMD Epyc server, we've noticed that is impossible to take a successful backup through PBS. We can just do a local backup or maybe over a NFS storage, but not over PBS datastore.
When we try making a backup over PBS, it seems to start correctly but then it hangs in this way:
INFO: starting new backup job: vzdump 806 --remove 0 --mode snapshot --node n9a --storage pbs1
INFO: Starting Backup of VM 806 (qemu)
INFO: Backup started at 2021-11-18 22:28:22
INFO: status = running
INFO: VM Name: test-debian10
INFO: include disk 'scsi0' 'local:806/vm-806-disk-0.qcow2' 200G
INFO: backup mode: snapshot
INFO: ionice priority: 7
INFO: snapshots found (not included into backup)
INFO: creating Proxmox Backup Server archive 'vm/806/2021-11-18T21:28:22Z'
INFO: issuing guest-agent 'fs-freeze' command
INFO: issuing guest-agent 'fs-thaw' command
INFO: started backup task '3c770816-0edc-4107-9df1-221cb2c676b1'
INFO: resuming VM again
INFO: scsi0: dirty-bitmap status: created new
INFO: 0% (408.0 MiB of 200.0 GiB) in 3s, read: 136.0 MiB/s, write: 10.7 MiB/s
And stays in that way for a long time (more than 30 mins) and then it hangs in timeout... server are connected with each other with 10 GbE network and everything works well. I can even restore a backup from PBS to the new server and it works perfectly!
I cannot just do backup over PBS.
In also, with Linux VM, when it hangs in this way, then VM filesystem's got dirty and I must reboot and do fsck /dev/sda1 manually during boot!
No one knows how to fix this?
Thanks in advance...
Paolo
