My company uses Veeam Backup & Replication to make backups of our VMWare-hosted virtual machines. We use Veeam's SQL log backup feature for our MS SQL 2008 R2 server. Our backup repository resides on a Buffalo TeraStation TS-XLBF NAS.
At some point, we started getting error reports of this kind:
Failed to save transaction log backup file for database [DB name] to repository: Der angegebene Netzwerkname ist nicht mehr verfügbar. Error code: 64 Failed to flush file buffers. File: [path to .vlb file]. Failed to backup file. File: [file description] Failed to save transaction log backup file for database to repository: Der angegebene Netzwerkname ist nicht mehr verfügbar. Error code: 64 Failed to flush file buffers. File: [file name]. Failed to backup file. File: [file description] Transaction logs from some databases were not backed up for the last 38 intervals
The text in German (Der angegebene Netzwerkname ist nicht mehr verfügbar) states, The specified network name is no longer available.
As it turns out, this error occurs every Monday night during the scheduled time for the regular backup jobs. The VM backups themselves are completed successfully, only the SQL log backup is affected.
The error messages are sent every few minutes but stop once we restart the system that's running Veeam. Looking for the cause of the issue, I found an article on the Veeam website describing this exact error message. The article identifies issues with de-duplication on NTFS storage as the root cause. Since the TeraStation is a Linux system using XFS, which doesn't support de-duplication, I don't think this applies in our case.
I also tried accessing the backup repository using the path mentioned in the error message, while the error messages were still being generated. I was able to access the resource successfully, so I guess name resolution is not the issue.
The event logs on the machine running Veeam don't show any error messages. The regular backups of all VMs run without errors.
What could be the cause of the failing SQL server log backup jobs? Is there a way to gather more information about the failure?