>  The server will attempt to use an empty volume
> whenever it must continue a file to another tape.  Therefore, I would
> expect use of the filling volumes to occur over an extended period of
> time, depending on the size distribution of your files.

If I read this correctly, then what will happen is that each migration will
start using a tape that is in "filling" status, and when it is full, and the
file being written is split over the end of a tape, then a new scratch mount
will be requested.  Since most of the "Q Contents" lists I have seen have
partial files as the last block, then indeed the number of "filling" tapes
will drop slowly.

I think it is time to quit thinking about this and "just do it" - we are
planning on turning off colocation tomorrow.  I have a Rexx program that
counts the tapes in each category (full and filling) as well as total numbers,
so between that and some activity log searched for tapes added and deleted, we
should be able to track the net effect of all this.

Thanks again for the assistance.

Jerry Lawson
jlawson AT itthartford DOT com


________________________Forward Header________________________
Author: INTERNET.OWNERAD
Subject: Collocation Concern
02-12-96 01:01 PM

>Thanks for the information - Your explaination was very clear.  My appreensio
>ns are not from the design, but rather the result of our changing directon
>when we have a large number (280) of "filling" tapes.  I have visions ofthe
>first migrations after we turn colocation off having many extra mounts bcause
>these tapes are close to being full.

You've got a legitimate concern, but I doubt the problem will be as
bad as you might think.  The server will attempt to use an empty volume
whenever it must continue a file to another tape.  Therefore, I would
expect use of the filling volumes to occur over an extended period of
time, depending on the size distribution of your files.  This means that
you probably will not see an enormous surge in mount requests during
your first few migrations.  The downside of this is that the number of
filling volumes may not be reduced as quickly as you would like.

Dave Cannon
ADSM Development