Cluster vms fail randomly with "no such host is known"

[DanielJ]

I have seen this here and there before, and it always went away by just retrying the job. Now I have a new VBR server where I'm trying to back up a Hyper-V cluster previously backed up by an older VBR server, and now I get this error around 50% of the time. Nothing changes between attempts, on-host proxy servers, the job even starts reading disks and all seems fine, before suddenly cancelling everything with "Error: resolve: no such host is known". It sounds like dns issues, but there is no clue exactly which server is trying to look up what other server. I made sure from the start that the VBR server could look up the hosts, forward and reverse, but it seems to have made no difference. What I really don't understand is how it can act so randomly, and how VBR can get to the point of starting to read data (and write it to the repository) before deciding that something is wrong. I have a ticket going on but so far we have nothing. Maybe we will get somewhere soon, but in the meantime, does anyone else have the same Hyper-V problems constantly reoccurring? I have searched and only found it with some replication job.

Ticket ID: 04198972

[foggy]

Hi Daniel, I can see that the case has just been opened today, please allow some time for deeper logs analysis by our engineers. The behavior looks strange, indeed.

[DanielJ]

I learned from support that this error can be just about any networking issue and has likely nothing to do with dns in this case. Instead it was probably caused by connections being dropped due to high load on network and repository. Much more understandable than what the error message indicated. All seems fine now that most full backups have completed.