I have a back up job that never had any failures, but took very long time to complete because 2 proxies configured for the job were hard set to user network transport mode. A few weeks ago I switched to HotAdd mode and saw a significant performance increase as expected, but random VMs started to fail, not expected.
What I noticed is random VMs (1 or 2 out of 20) would fail on a first run, but would be successful on the 1st retry.
The error message would be consistent for any failed vm
This is QNAP ts-451u with Frimware - 5.1.1.2491. QNAP Antivirus is Off, so that's ruled out.11/6/2023 6:37:00 PM :: Processing SERVER-NAME Error: The process cannot access the file because it is being used by another process.
Failed to create or open file [\\qnap-451u\vmbackup\another_folder\SERVER-NAME.vm-718980D2023-11-07T105020_0AC4.vbk].
File '\\qnap-451u\vmbackup\another_folder\SERVER-NAME.vm-718980D2023-11-07T105020_0AC4.vbk' locked by 0 processes:
Failed to open storage for read/write access. Storage: [\\qnap-451u\vmbackup\another_folder\SERVER-NAME.vm-718980D2023-11-07T105020_0AC4.vbk].
Veeam Support suggested switching to per VM metadata files. I went ahead and cloned the existing job and disabled the old one.
Running a new job has made things worse, now instead of 1 or 2 VMs failing on a first run I have 14 - 16 VMs failing on a first run, but with each retry the number of failed VMs gets lower and lower.
As you can tell from the error message we are backing up to an SMB share. AV exclusions are configured on all VBR and VBR Proxy servers.
Case: 06392979