In addition to the considerations mentioned by Tim, there is an additional
advantage to using migration if you are transferring data from a disk storage
pool to a collocated storage pool.  If this is done using Move Data,
multiple passes are made through the database to ensure that files are
written to the target pool in a collocated manner.  This additional database
activity is not necessary with migration, because file information is
stored in the database in such a way as to optimize migration processing by
node.

Dave Cannon
ADSM Development
------------------------------- Referenced Note ---------------------------
Debra,
Debra,
        I would think the performance would be similar as they're both
performing the same function.  The advantage to using migration is that
you can kick off multiple concurrent migration processes and you can
also control when they run by lowering and raising the high migration
threshold for your disk storage pool(s) via the ADSM admin scheduler
function.  The average size of the files you're migrating from disk to
tape will have a lot to do with what kind of thruput you'll see.  I run
5 concurrent migration processes every morning on our MVS system - it
usually takes about 3 to 4 hours to move 18-20 GB so we're seeing
thruput of about 5-6 GB per hour or about 1 GB per hour per migration
process.   You should also take a look at a following parameters in your
server options file - changing these can make a big difference -

MOVEBATCHSIZE
MOVESIZETHRESH

Tim Pittson
tpittson AT himail.hcc DOT com

>----------
>From:  Debra Jeanette Young[SMTP:djyoung AT VOYAGER DOT NET]
>Sent:  Tuesday, July 02, 1996 5:26 PM
>To:    Multiple recipients of list ADSM-L
>Subject:       Benchmarks for MOVE DATA vs MIGRATION
>
>I am looking for information for using MIGRATION vs MOVE DATA on MVS
>storage
>pools.  Our current MOVE DATA takes about 35 minutes to move about
>915M.  I
>want to convert this to migration and was curious if there would be a
>significant increase in elapsed time.
>Deb
>