Q: TimeMachine errors when backing up to XSan Volume

Hi all,
I've been trying to solve this problem for a while. I have an apple xraid unit with a timemachine enabled afp share that I want to use for client backups over our network.

10.5 and 10.6 clients initially can configure timemachine preferences to use this share and the first backup image gets written. However, on subsequent backups a new sparse-image bundle is created with a numerically incremented filename. (i.e. client_backup.sparsebundle gets created then sometime later timemachine creates client_backup(01).sparsebundle).

Also, I've seen subsequent backup attempts simply fail with "Volume cannot be accessed (exit code 109)" errors. From the client console, I see these log messages:

Mounted network destination using URL: afp://client_backup@servername.local/TimeMachine
com.apple.launchd[1] (com.apple.netauth.sysagent[1333]) Exited with exit code: 255
diskarbitrationd[13]: unable to repair /dev/disk2s2 (status code 0x00000008)
om.apple.backupd[1332]: Failed to mount disk image /Volumes/TimeMachine/client_backup.sparsebundle (DIHLDiskImageAttach returned: 109)

Has anyone seen similar behavior? I'm stumped as to why the first backup is apparently successful but subsequent backups fail. Thanks for reading.