Does "media multiplexing" take precedence over "Prioritized" in a Storage Unit Group?

[sclind]

I have two tape silos attached to one backup master.  I prefer that each silo be used by the Business Unit associated with that silo.  But in case one silo goes down I allow the backups to run on the other silo.  (Better to have the backups float to the other silo than not run at all).

So I have defined two storage groups - each has both silos defined (in opposite order of course). The Storage Unit Selection is set to "Prioritized".  So I would expect Business Unit A's backup to normally run on silo A and Business Unit B's backups to normally run on Silo B.

But somewhat often I see 'A' jobs using the 'B' silo or 'B' jobs using the 'A' silo.  My only guess is that Netbackup tries to Multiplex to an existing drive if it can - even if that drive is on the lower priority Storage Unit.

Does anyone know if this is what Netbackup is doing?

[RamNagalla]

since you are looking for Silo unavaliable situation.. you need to select the option Failover in storge unit group instead of Prioritized, 

because Prioritized will choose the next STU even the 1st storge unit is busy.. ( not down case..) however  if you select Failover it will chose the next storage unit only when the 1st one is down or out of media..

so goahead and select Failover option in Storge unit group and see how it goes..

[sclind]

It looks like the only difference between these two is "not busy".  I guess if all my drives were busy in one silo I would want the job to use the other silo. (For example, we run some database archive log backups that can't afford to wait for a drive).

So Prioritized still seems like the correct choice (for me).

(When I have seen the situation I described occur the storage unit still had available drives).