>"MC Matt Cooper (2838)" <Matt.Cooper AT AMGREETINGS DOT COM>
>>Does anyone have a script that will arrange/control the submission of
>>multiple restore sessions to avoid having sessions having to wait for
>>other sessions to finish using a tape?

From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On 
Behalf Of Dale Jolliff
>I'm probably just missing the obvious, but why not just collocate by
>filespace?
>Then you can do all the restore scripting you want with no worries.

This is an interesting question, and goes to the heart of disaster
recovery priorities. In the best of all possible (restoration) worlds,
each filespace for each TSM client is collocated for all primary and
copy pools. Naturally, this condition would create a library resource
nightmare. Still, it's a possibility for some, particularly if offsite
electronic vaulting, disk-based storage pools, or virtual tape systems
(VTS) are in the mix.

A compromise would be collocation by filesystem only for those servers
deemed critical. In my experience, I've not often seen multiple requests
for the same tape volume hold up restores *that* badly (unless we're
talking about large database files, in which case they should probably
be in their own primary/copy storage pools anyway). Any system whose SLA
merits the sort of fine-toothed combing that Matt's goal requires
probably needs to be the subject of some sort of real-time replication,
rather than large-scale file restoration in order to meet the SLA.

(Random musings came up with an odd idea that would "fix" multiple
requests for the same tape--some sort of anti-collocation that would
scatter files as far apart of possible.)

--
Mark Stapleton