I configured new and converted several existing clients from SMB to NFS repository. It is a simple scale-out repository with NFS extent on synology/qnap and S3 wasabi for offload. With one client I got error below. I opened support ticket and support tech tells me that NFS extents are not supported with object storage offload. Is there any truth to this? I have more similar setups with same hardware work without issues.
[06.03.2020 04:49:47] <22> Error WriteSynthetic is not supported in NfsStgFile (Veeam.Backup.Common.CCppComponentException)
[06.03.2020 04:49:47] <22> Error in c++: Failed to copy internal FIB data blocks to a new block store.
[06.03.2020 04:49:47] <22> Error in c++: Failed to list storage items (root folder: '').
[06.03.2020 04:49:47] <22> Error in c++: Failed to copy compacted block store.
[06.03.2020 04:49:47] <22> Error in c++: Failed to compact index data. Path: [nfs41://synology:/volume1/backups/ArchiveIndex/a944fd95-2712-4150-a345-adf7669f1210/f794817a-812c-e819-54c7-70c1d3eff765/index_data.vbk].
[06.03.2020 04:49:47] <22> Error in c++: Failed to process method 'ArchRepo.CompactIndex'
I switched storage back to SMB while I was waiting for response on ticket. It allowed offload to continue. I will change it to NFS again today.
Also this only happens to a single vm.
That helps. I also read the case notes. You are not doing anything wrong persay and it does appear to be a quirk. Support is going to escalate and work with you on a solution. They will reach out soon if they haven't already.
Dustin Albertson | Director of Product Management - Cloud & Applications | Veeam Product Management, Alliances
I've reviewed the case details. Currently, it's unclear what is causing this behavior so please keep working with the support engineers directly - they might request additional information. We'll keep an eye on the case flow.