ADSM-L

Re: Offsite Reclamation Problem

1999-01-21 17:20:18
Subject: Re: Offsite Reclamation Problem
From: Bill Colwell <bcolwell AT DRAPER DOT COM>
Date: Thu, 21 Jan 1999 17:20:18 -0500
In <7D56E0277E05D211971A000058011927027DFE97 AT msgmsp4.norwest DOT com>, on
01/21/99
   at 09:48 AM, "Glass, Peter" <Peter.K.Glass AT NORWEST DOT COM> said:

>Hello, ADSMers!
>Here's the scenario:
>Offsite tape reclamation happens to be still running when the scheduled
>command executes to change the access status of any OFFSITE tapes that are
>Onsite
>(i.e., upd v * acc=of wherestg=copy_off1 whereacc=readw,reado
>wherest=fil,ful).
>The new OFFSITE tape that is currently  being written to then dismounts, on
>account of its newly changed status, and the courier takes it away to its
>offsite location.
>The Offsite Reclamation process is hung at this point, the input tape stays
>mounted, and this state of affairs will remain until the Server is
>stopped/restarted.
>Why doesn't the Reclamation process simply ask for another scratch tape
>when the current output tape dismounts?
>Would this be a procedure problem-or a bug?

It's your call!  There's no penalty for calling IBM support and expressing
your opinion to them.  They will either agree with you or call it a feature,
at least that has been my experience.

In the meantime you need to fix your process.  You can add 'preview=yes' to
the upd v command, capture the output and generate specific update commands
leaving out the volume in use.  Or you can do something with the sql
interface such as

select left(volume_name,16), last_write_date from volumes where stgpool_name
= '9840ACOPY' and status ^= 'EMPTY'
and access ^= 'OFFSITE' >vol.lst

and

select * from processes >proc.lst

and massage the output.


>We never had this problem under V2.
>We are now running under MVS-OS/390 V3, Release 2.0.
>Any comments or suggestions would be most appreciated.
>Thanks, in advance.
>Peter Glass
>Norwest Services, Inc./Wells Fargo Company

--
-----------------------------------------------------------
-----------------------------------------------------------
Bill Colwell
Bill Colwell
C. S. Draper Lab
Cambridge, Ma.
bcolwell AT draper DOT com
-----------------------------------------------------------
=========================================================================
<Prev in Thread] Current Thread [Next in Thread>