I said that the largest user of data is moved first in disk to tape migration.
and for tape to tape this
is proving not to be true for me as I open up 3490 migration from 3480.

To wit:   I have 150+ 3480s that are: full, filling and maybe 1, empty.
I set NEXTPOOL to the 3490s.  Both have COLLOCATION ON.   I change the
Hi/Low migration numbers of the 3480 pool to be just 1 percentage point
below its current state.   ( BTW: I have *NOT* made 3480s read/only )
Changing the settings immediately triggers 3480 mounts for R/O and 3480 & 3490s
mounts ( current and scratch ) for writing.   When all's done I go back
and do a QUERY CONTENT on the newly written 3490 tape and I do QUERY OCC on all
my users and I get a mish-mash of results.  ( User's data partly on 3480&3490s)

I expected ADSM to pick my largest user and move it all to 3490 in toto.
It doesn't.    Can I get ADSM to move large users (only) to a new level?
    Any comments?  help?
          thanks .. joe.f.

On Wed, 26 Apr 1995 08:17:32 MST Dave Cannon said:
>For sequential-access storage pools, the high and low migration
>thresholds refer to the percentage of volumes in the pool that contain
>data.  If you have 100 volumes in a sequential storage pool and 69
>volumes contain data (it doesn't matter how much), then the
>migratable percentage is 69%.  In this situation, if you set
>high migration to 60%, migration should begin.
>
>When migrating from disk, the server selects the node with the
>largest filespace in the storage pool and then migrates ALL data
>for that node to the next storage pool.  Migration from disk proceeds
>in this manner, regardless of whether the next storage pool is
>collocated.
>
>Dave Cannon
>ADSM Development