Storage pool load balancing logic currently seems geared towards tape devices, but when it comes to disk devices, storage capacity needs to be taken into account.

Right now, my backup jobs use the same B2D device until it's completely full, and only start using a different device in the pool then.  That’s certainly better than failing the job, but things could be improved.   I have to accept the full error state of my most commonly-used  B2D device as a normal condition. This makes it harder to spot (i.e. easier to overlook) true error situations.

This is similar to the thread at

https://www-secure.symantec.com/connect/ideas/backup-exec-any-disk-storage-round-robin-device-selection

except that there are some simple job settings that could help drive storage device selection:

1.     The B2D devices' "threshold" values could be part of the equation. The devices with the "best" threshold status would be the first filter for candidate devices to put the backup data on.

2.     The device with the "most" available capacity could be the one selected.