I am running ADSM server V.2.1 level 13 on MVS. I intend to upgrade to V.3 
within
the next 2 months.

I am facing this reclamation problem. IBM claims it is working as intended ! I 
am
pretty curious about how V.3 is handling this.
Reclaiming tapes in the offsite copy storage pools is a real pain.
The offsite copy storage pools are using 3490 tape units with stackers.
Collocation is on for these. The primary storage pools are also 3490s in an STK
tower. For these, collocation is done at filespace level.

At the time the reclamation process is kicked off, all the tapes are marked
offsite. It is running pretty well in the beginning - when it is only asking for
scratch tapes. Then it starts asking for tapes that it have already had mounted
ealier in the process. And it might ask for the same tape several times during
that process.
Since the tape unit stackers are only loaded with scratch tapes, it requires a 
lot
of human intervention to manually mount these specific tapes.
Why cant ADSM just finish writing a tape once it is mounted ?

Occasionally this also happens during storage pool backup.

The way transactions are batched within ADSM is probably partly the reason for
this. However, if these transactions were ordered in accordance to collocation
rules, it might eliminate this problem.

I have not researched how often this occurs for the primary storage pools in the
tower.


Regards

Carsten Moldrup
CSC Computer Management
Copenhagen
Denamrk