ADSM-L

Re: PMR 02528, 082

2003-02-07 11:16:07
Subject: Re: PMR 02528, 082
From: Alex Paschal <AlexPaschal AT FREIGHTLINER DOT COM>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Fri, 7 Feb 2003 08:14:24 -0800
Bill,

The real question is, "what is your client data profile?"  While I don't
think I'd mirror a stgpool used solely for filesystem backups, I myself
would be inclined to mirror a stgpool that contains "live data" such as
archived txn logs from DB2 which then get reused or fileserver cleanup
archive-delete processes.  If, after either of the aformentioned data gets
sent to disk, the disk fails before your daily (bi- or tri-daily? Somewhere
in 8-24 hours, anyway) migration/backup, you could have a heck of an
exposure.

For example, say, your DB2 backed up last night, and does more or less
hourly txn log archives so that you can roll forward to the last instant
(important for financial databases like SAP, so a failure would be very high
profile).  You backup stg and migrate disk to tape once in the morning, once
in the evening. Your log archives at 5:30AM, gets moved to tape.  DB2
archives again at 6:30, 8:00, 9:30, 10:30.  However, after the 8:00 archive,
the disk crashed and the volume went offline to TSM.  Your next archives
still succeed because they go to another disk volume or to nextstgpool.
HOWEVER, they'll be useless.  You can't meet your SLA and recover to point
in time after 6:30AM because all your txn logs build off of the 8:00 one
that is lost.  Too bad you didn't mirror.

2nd example, say Accounting runs monthly financial reports.  Your SLA says
there's a 6 month retention on those reports, then they get archived (7
years, in case of audit) and deleted to save disk space on the fileserver.
6 months later, an automated script (because this is a well-defined process)
goes to archive them and delete them from disk with an archive
-deletefiles=yes.  An hour later, before migration or backup stg happens,
the disk crashes.  Are you going to remember to restore from backup so that
you can re-archive them, or is that data lost?

Hope this helps you decide whether you want to mirror or not.

Alex Paschal
Freightliner, LLC
(503) 745-6850 phone/vmail


-----Original Message-----
From: William Rosette [mailto:Bill_Rosette AT PAPAJOHNS DOT COM]
Sent: Friday, February 07, 2003 6:19 AM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: PMR 02528, 082


Does any TSM gurus have any suggestions for our AIX admin

Thank You,
Bill Rosette
Data Center/IS/Papa Johns International
WWJD
----- Forwarded by Bill Rosette/IS/Corporate/Papa Johns on 02/07/2003 09:20
AM -----

                      Steven Schraer
                                               To:       hulls AT us.ibm DOT com
                      02/06/2003 05:09         cc:       Bill
Rosette/IS/Corporate/Papa Johns
                      PM                       Subject:  PMR 02528, 082





We are having some problems with the first link below.  The second is
working.

I have been looking at the Redbook.  Gettting started with Tivoli Storaage
Manger: Implementation Guide.

It looks like for storage pools are recommend for raid 0 (mirroring), raid
0+1 (mirroring & stripping) or raid 5 (distributed parity).  These are safe
methods to protect the storage pool.  Do you know of any companies that use
just raid 1 (stripping) on their storage pools?  Is there an issue of tsm
loosing a storage pool and the database having issues due to the lost
storage pool data?

Also there is no suggested raid for the tsm database.  Since we mirror the
tsm database with the tsm software, can we use raid 1 (stripping) so that
we can get a performance increase on accessing the database?  Or should we
just create the volume group with no raiding but lay the database over
multiple smaller disk (i.e. use eight 18.2 Gig disk instead of 4 36.4 Gig
disk).  What will give us the best performance.

Can I get an answer on this tomorrow morning.  I have a meeting to discuss
the performance issues tomorrow (Feb 7, 2002).

Thanks for your assistance!

Steven M. Schraer
502-261-4148
----- Forwarded by Steven Schraer/IS/Corporate/Papa Johns on 02/06/2003
04:57 PM -----

                      Bill Rosette
                                               To:       Steven
Schraer/IS/Corporate/Papa Johns@Papa Johns
                      02/06/2003 03:51         cc:
                      PM                       Subject:  PMR 02528, 082





Could not get into the link below.

Thank You,
Bill Rosette
Data Center/IS/Papa Johns International
WWJD
----- Forwarded by Bill Rosette/IS/Corporate/Papa Johns on 02/06/2003 03:53
PM -----

                      Stephen Hull
                      <hulls AT us.ibm DOT com        To:
bill_rosette AT papajohns DOT com
                      >                        cc:
                                               Subject:  PMR
                      02/06/2003 03:36
                      PM






Hi Bill,
     Here is a link to the Performance Tuning Guide on the web:
ftp://w3.ibm.com/support/ats/documents/TSM42PDG.pdf

Regards,
Stephen Hull
IBM Tivoli Storage Manager Support

More questions?  Search our new TSM KnowledgeBase at:
http://www-3.ibm.com/software/sysmgmt/products/support/IBMTivoliStorageManag
er.html

<Prev in Thread] Current Thread [Next in Thread>