ADSM-L

Re: Storage pool DSNs

1998-12-15 10:13:20
Subject: Re: Storage pool DSNs
From: Steve Fletcher <sfletcher AT GIANTOFMARYLAND DOT COM>
Date: Tue, 15 Dec 1998 10:13:20 -0500
Thomas,
When you define your sequential storage pools you assign a device class.
In order to use the CA-1 TMS/VMS you will need to define two device
classes.  It is there you control the DSN assigned to the tape dataset.
Unfortunately you are limited to eight (8) characters.  We use ADSM.BKP for
onsite tapes and ADSM.OFF for offsite tapes.  ADSM appends additional
information to these first eight characters.  (.BFS for backups/archives
and .DBB for database backups).  You then tell CA-1's vault management to
look for the appropriate HLQ.

If you have tapes assigned to a storage pool/device class then you should
create a new offsite device class and storage pool.  If you change an
existing device class the tapes already defined will not be accessible. (as
per IBM ADSM documentation).

If you have any additional questions you can contact me directly.

Note: My answer is based on ADSM/MVS version 2.  I don't know if this has
changed under version 3.




Please respond to "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>

To:   ADSM-L AT VM.MARIST DOT EDU
cc:    (bcc: Steve Fletcher/GIANTMD/US/Ahold)
Subject:  Storage pool DSNs




My site is in the process of installing ADSM under MVS. We plan to have all
copy storage pool volumes written on a group of 3590 tape drives. We will
probably have volumes for at least some of the primary storage pools
written
on the same drives. As I understand matters, the datasets on storage pool
volumes written on the 3590 drives will all have the same DSN, regardless
of
which storage pool is involved and regardless of whether it is a primary
pool
or a copy pool.

My site uses the Vault Management System feature of CA-1 to control the
movement of tapes stored offsite. VMS can identify tape datasets that are
candidates for offsite vaulting by DSN, job name, or a combination of the
two.
I would assume that all tapes written by ADSM server functions will share
the
same job name.

Have I missed something, or has IBM neatly and systematically precluded
every
possible means of having VMS identify ADSM tapes that should be sent
offsite?
<Prev in Thread] Current Thread [Next in Thread>