If you get any i/o problems with the library due to a hardware problem all the
scratch tapes can get pulled into one storage pool, get a little data written
to them, then marked as read-only. If you have or more tapes suddenly used
up, you'll be left with a lot of clean up. Like manually
Managing scratch volumes in multiple libraries AND also on stg pool level
is just too much work for little gain.
Like Wanda said before the stg pool parm works great in small setups with
limited number of tape slots.
With multiple tsm servers each having multiple stg pools and connected to
We just went through this, having a high percentage of VTL volumes marked FULL
when they only contained a few MB, when power maintenance in the data center
went wrong. We were able to define new volumes, fortunately, but 'move data'
jobs don't run that fast actually. If maxscratch had been set