ADSM-L

Re: Controlling BFS Deletes after EXPIRE

1997-02-28 17:26:19
Subject: Re: Controlling BFS Deletes after EXPIRE
From: Hans Beierle <hbeier01 AT DE.EDS DOT COM>
Date: Fri, 28 Feb 1997 14:26:19 -0800
Paul,

we noticed a similiar problem, which started to happen since ADSM/MVS server 
level 2.1.10. The only difference to you is that our 'BFS' files
are tape volume datasets. EXPIRE INVENTORY started to work as expected but then 
all the sudden, ADSM CPU usage bumped up. You couldn't see any
further progress in the EXPIRE process and virtually all other tasks under ADSM 
were put on hold. We reported the problem to IBM and send
them a dump and additional information some time ago, but the trouble ticket is 
still open.
Just recently we noticed some strange behaviour during EXPIRE regarding the 
volume history file. We saw quite a lot of I/O activity and
mainly a permanent allocation of this file, which we considered unusual. We 
kept stg entries for about 3 weeks, so storing about 7000 stg
entries. Due to this observation, we changed the stg entry retention to 7 days 
(Del volh tod=-7 Type=STGN/STGD/STGR). This reduced the number
of stg entries by about 50% AND it obviously solved our problem. Since about 2 
weeks, it never happened again.

Hans Beierle
--
/----------------------------------------------------------------\
/----------------------------------------------------------------\
| Hans Beierle                      Internet: hbeier AT eds DOT de      |
| EDS Informationstechnologie       Voice: ++49 6142 803170      |
| und Service (Deutschland) GmbH    FAX:   ++49 6142 803030      |
| Eisenstr. 56 (N15)                                             |
| D-65428 Ruesselsheim, Germany                                  |
\----------------------------------------------------------------/


Paul D Brown wrote:
>
>   Dwight,  the BFS files I referred to, the flat files in my second level
>   storage group that is sometimes
>   tape, but in my case I use only DASD.  The BFS in the file name is an IBM
>   default, I believe.
>   I have to run EXPIRE every day just to keep the storage groups clean.
>   The ADSM region
>   goes into a high CPU usage state then hangs.  This happens everyday,
>   sometimes after
>   EXPIRE ends and sometimes right before it completes.  We have to cancel
>   ADSM because
>   of the high CPU usage and on the reload the clean up from the EXPIRE
>   begins.  The BFS files
>   are mounted one by one to be deleted.  The delay to get signed back on to
>   ADSM to do some
>   restores was a record 7 HOURS yesterday.  Users are starting to complain
>   as you would expect.
>   As I stated all of my second level storage group is kept on disk, which I
>   would expect to help
>   speed things up.  This storage group is what the BACKUPPOOL  storage
>   group is migrated too.
>   Hope this helps.  I need some help.
>   Thanks,
>   Paul Brown
>   PDBrown.Ashland.com
>   606-357-7585
>   To:  ADSM-L @ VM.MARIST.EDU
>   cc:        (bcc: Paul D Brown)
>   From:     decook @ AMOCO.COM
>   Date:     02-26-97 08:22:09 PM CST
>   Subject:  Re: Controlling BFS Deletes after EXPIRE
>
>     ---------------------------------------------------------------
>
>        Paul, what is being referred to by BFS ? (backup file system comes
>        to mind but I hate guess work...)  Our MVS test server I believe is
>        at 2.1.12 I can't say right now cause it isn't running... but this
>        sounds like something I need to look into...  the only hits I got in
>        my manuals for BFS was an arbitrary name  used in conjunction with
>        data set prefixes on tapes...  Could  you enlighten me a little
>        more... Currently it sounds like when it died it left something in
>        an "easy to tell something went wrong" state and I don't think I'd
>        want ADSM to allow access until it had straightened itself out...
>        but I need a
>        little more info...
>        later
>            Dwight
>
>   ______________________________ Reply Separator 
> _______________________________
>   __
>   Subject: Controlling BFS Deletes after EXPIRE
>   Author:  ADSM-L at unix,mime/DD.RFC-822=ADSM-L AT VM.MARIST DOT EDU
>   Date:    2/26/97 7:56 AM
>
>     Hi ADSMers,
>       I am running ADSM/MVS server 2.1.12.   I am having trouble with ADSM
>     hanging up after EXPIRE
>     ends.  After a cancel and restart of the ADSM region The BFS deletes
>     begin as cleanup from the EXPIRE.
>     The ADSM server will not allow access until all deletes are complete
>     which normally is 4 to 5 hours of
>     down time.  My question, Is there any way to prevent the EXPIRE clean
>     up from starting after the ADSM reload?
>
>     Thanks,
>     Paul Brown
>     Storage System Manager
>
>     PDBROWN.ASHLAND.COM
>     Ashland Inc.
>     Lexington, Kentucky
<Prev in Thread] Current Thread [Next in Thread>