Replication failure improvement - staggering or bandwidth limit

[Elliott Partridge]

I routinely get replication job failures between the hosts in my 2-node Proxmox cluster. I have grown accustomed to ignoring them as they always recover on the next run. Is there a way to avoid these failures? I'm thinking of staggering the replication schedule between the two hosts, as they currently run jobs at the same time. Another thought would be to limit bandwidth of replication, but not sure that I want to slow down everything like that.

Has anyone had experience with this and care to share?

 

[Philipp Hufnagl]

Hello

Could you give me your journal of the last 2 days (assuming the issue occured in this time)?

journalctl --since '2023-10-22' > $(hostname)-journal.txt

Also, please share your /etc/pve/storage.cfg

 

[Elliott Partridge]

Hello

Could you give me your journal of the last 2 days (assuming the issue occured in this time)?

journalctl --since '2023-10-22' > $(hostname)-journal.txt

Also, please share your /etc/pve/storage.cfg

Thanks for the prompt response!

I've attached the requested files, zipped up. There are many ZFS operations resulting in failed: got timeout, mostly if not exclusively occurring during backup jobs.