When using BackupVSS to create consistent Backups of Notes Databases during a snapshot backup BackupVSS operation, only Databases "physically" lokated in DominoDataDir will be honored.
Databases lokated outside of DominoDataDir f.e. on another disk, mounted via a directory link (xxx.dir) will generate an error and will not be freezed for VSS Backup.
Note: As we should know, BackupVSS uses features from DominoBackup to bring Databases into a consistant state. When BackupVSS triggers DominoBackup the first time, it creates DominoBackup.nsf if not present, using default configuration settings.
The second time BackupVSS triggers DominoBackup errors notifications will be send to LocalDomainAdmins by default. This will occure even if DominoBackup.nsf is not actively used !
Sure you could simply disable error notification in DominoBackup.nsf to avoid the notifications. Especially it the snapshot backup opperation is marked as successfully.
Unfortunately, Databases lokated on another disk will potentioally not be usable after restore, because BackupVSS doesnt care for them to be consistent during backup.
The first time the backup servertask runs, it creates dominobackup.nsf database, which holds the configuration.
Therefore the first backup is not excuting any backup.
Domino backup and also Domino VSS Backup requires a configuration!
backupvss is a VSS Writer implementation. When the task is started, it registers Domino as a VSS writer waiting for a snapshot operation to occur triggered by a backup application (like Veeam or others). backupvss starts the Domino backup servertask, which then performs the freeze and backup operations to create the inventory.
Only supporting one volume is a design decision early in the project. Because it is a flexible integration with many different ways to configure a backup integration and the complexity snapshot and restore adds, it's not very likely multiple volume support is added soon.
The snapshot is only needed for the NSF volume. NIFNSF, FT, Translog and DAOS should be a separate volume. NSF should then fix on a single data volume.
The error messages for directory links are on purpose to prevent data loss. The team made sure the physical location is checked and a proper error is returned.
But let's assume for a second we would add multiple volume support to explain why this is difficult to implement.
Your directory or database link would point to a different volume. When restoring this database the right volume and point in the volume would need to be found and that volume must have been mounted.
There are many edge cases like this when implementing a restore for a snapshot based backup.