The ENOSPC error means that that temporary snapshot space has filled up. By default, Veeam reserves a very low limit for this purpose; if during the backup there are many disk writes, that small space fills up in seconds and the backup fails.
I think the solution to this problem is to increase the snapshot size. Take a look at this doc:
The likely cause is that the snapshot storage does not grow fast enough: by default, it grows in portions of 1 GiB, and if the server writes data faster than new portions are added, the snapshot overflows (error 28). The size of these portions is set by diffStorageMinimum in the [blksnap] section of /etc/veeam/veeam.ini. Its value is in 512-byte sectors, and the default, 2097152, is 1 GiB. Could you please double it to reach 2 GiB?
[blksnap]
diffStorageMinimum = 4194304
The kernel module only reads this value when it is loaded, and it stays loaded between backups, so restarting veeamservice is not enough. While no backup is running, you can restart the service and unload the module:
The diffStorageMinimum value was increased from 2097152 to 4194304 in /etc/veeam/veeam.ini, and the new setting was applied to the running kernel module.
After several days of snapshot overflow errors, the backup finally completed successfully without any issues.