ADSM-L

Re: Slow Offsite Reclamation (Was Tape drive recomendations)

2002-11-01 18:25:10
Subject: Re: Slow Offsite Reclamation (Was Tape drive recomendations)
From: Tab Trepagnier <Tab.Trepagnier AT LAITRAM DOT COM>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Fri, 1 Nov 2002 17:19:25 -0600
Tim,

I guarantee that the problem still exists in 4.1.5.0.

After seeing my TSM system take 12 hours to reclaim 100 MB of copypool
data I used that APAR as a reference and changed my directory disk pool to
FILE volumes.  The performance improvement was dramatic.  I use DLT 8000
for copypool, and when the source tape is LTO or 3575XL, I'm seeing the
full 8 MB/s capacity of the DLTs.  When reclaiming directories, the MB/s
throughput is still fairly low, but the object throughput  is on the order
of hundreds / sec.

Tab Trepagnier
TSM Administrator
Laitram Corporation









"Rushforth, Tim" <TRushfor AT CITY.WINNIPEG.MB DOT CA>
Sent by: "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
11/01/2002 09:25 AM
Please respond to "ADSM: Dist Stor Manager"


        To:     ADSM-L AT VM.MARIST DOT EDU
        cc:
        Subject:        Re: Slow Offsite Reclamation (Was Tape drive 
recomendations)


Yes! An old APAR cover's this issue, see below.  When you are running
offsite reclamation watch the throughput when it is reading from disk as
opposed to reading from tape.  It just crawls.

Apar IC15925:
--------------------------------------------------------------------
ERROR DESCRIPTION
During an offsite reclamation process; where files are being transfered
from
disk to tape.  ADSM fails to group the files together into transaction
groups.  Instead the files are processed sequentially.  Hence a file will
be
reclaimed and committed before the next file will be process. ADSM needs
to
honor the movebatchsize and movesizethresh options in the server options
file, so that the files will be grouped together properly for reclimation.
As a note, this problem will only be seen where off-site reclamation is
occuring from disk to tape.
COMMENTS:
After taking a very close look at the solution for this APAR, it was
determined that the code needed for this performance enhancement is
significant because it requires a restructure in the offsite reclamation
transaction processing.  A requirement can be taken to include this
performance improvement in the next release or version.
--------------------------------------------------------------------

Tim Rushforth
City of Winnipeg

-----Original Message-----
From: Steve Schaub [mailto:Steve.Schaub AT HAWORTH DOT COM]
Sent: October 31, 2002 11:15 AM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: Tape drive recomendations

If I'm reading this correctly, only migrating my diskpool down to 10% is
hurting my DRM reclamation?
Are you saying that putting in a 10TB diskpool that never goes to
primary tape (only dr copies) would give me monstrously bad
reclamations?  Our current main diskpool is 100gb on SSA and has caching
on.  I haven't had a terrible problem with DR reclamation, though the
storage pool backup to DRM only has a throughput of about 5mb/sec.

-----Original Message-----
From: TRushfor AT CITY.WINNIPEG.MB DOT CA
Sent: Thursday, October 31, 2002 11:03 AM
To: CITY.WINNIPEG.MB.CA.TRushfor; VM.MARIST.EDU;.ADSM-L
Subject: Re: Tape drive recomendations


And the issue is still there if you don't use CACHE=YES but don't
completely clear your backup pool.

We have more disk in our storagepool than is required for one night's
incremental - so we thought keeping some backups on disk was a good
thing (why migrate to tape if you don't need the space).


Tim Rushforth
City of Winnipeg

-----Original Message-----
From: Bill Boyer [mailto:bill.boyer AT VERIZON DOT NET]
Sent: October 31, 2002 9:01 AM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: Tape drive recomendations

Be careful of your copypool reclamations with the disk cache turned on!!
There is a BIG performance hit on reclamation when the primary copy of
the file is on a DISK direct access storage pool. Then the
MOVESIZETHRESH and MOVEBATCHSIZE values are thrown out the window and
the files are processed one at a time.

What I've done to relieve the restore times is to not MIGRATE the disk
pools until the end of the day. That way restoring from last night is
quick. I had a client where they wanted CACHE=YES on a 60GB disk pool.
The offsite copypool reclamation ran for 2-days! Changed it so that
migration started at 5:00pm and nobody complained about restore times.

Bill Boyer
DSS, Inc.