ADSM-L

Re: [ADSM-L] *EXTERNAL* Re: TSM DB spaces

2017-01-09 11:14:11
Subject: Re: [ADSM-L] *EXTERNAL* Re: TSM DB spaces
From: "Rhodes, Richard L." <rrhodes AT FIRSTENERGYCORP DOT COM>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Mon, 9 Jan 2017 16:09:02 +0000
Another consideration is the number if AIX hdisks the DB2 filespaces are using. 
 AIX uses I/O queues for each hdisk, and not having enough hdisks can cause I/O 
problems.

Our biggest DB2 db (600GB) (non-dudup'ed) currently uses 4 db2 filespaces.  
These reside on 8 hdisks which are pooled into one vg.  Filespace logical 
volumes are spread across the vg hdisks using a 32mb PP size.  They run through 
two FCS adapters that are set to round-robin.    

https://www-03.ibm.com/support/techdocs/atsmastr.nsf/5cb5ed706d254a8186256c71006d2e0a/d1f54f4cd1431d5a8625785000529663/$FILE/AIX-VIOS_DiskAndAdapterQueueTuningV1.2.pdf


In other words:
  DB2 will spread I'O's across filespaces
  AIX/LVM will spread I/O's across hdisks
  MPIO will spread I/O's across FCS adapters
  hdisks should have paths to multiple storage front end ports
  storage will spread I/O's across all the backend physical disks

All of these spots can become bottlenecks.

This is how we setup all our Oracle databases, from the tiny to the +30TB. The 
details vary, but the idea is the same.  


Isn't storage such fun!!!!!!

Rick

 



-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf Of 
Sasa Drnjevic
Sent: Friday, January 06, 2017 11:16 AM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: *EXTERNAL* Re: TSM DB spaces

On 2017-01-06 16:54, David Ehresman wrote:
> I am running TSM 7.1.5 on AIX 7.1.  My TSM DB is currently made up of
> five 120GB filespaces.  I need to dramatically increase the size of
> the DB as we prepare for dedup.  I know I can have 128 DB filespaces.
> What, if any, are the performance implications of just adding more
> filespaces vs staying with a smaller number of much larger
> filespaces?  The filespaces are SAN attached XIV with the data spread
> over 180 drives.
>
> David Ehresman
>

From experience, I believe it all comes down to IOPS.

So, it all depends on how are your TSM DBspaces distributed over as many
as possible drives, SAS or SATA HDDs, or SDDs, RAID type and of course
the type and throughput of (Fibre Channel?) SAN.

Your case with 180 drives sounds very good to me...

I've never had more than four TSM DB spaces, but I always distributed
them well over FC SAN, some on HDDs, and some on pure SSDs (Storwize
V7000), some only on RAID6, some only on RAID10. The sizes at the moment
are 400 GB, 200 GB and 800 GB TSM DBs.

Hope it helps...

--
Sasa Drnjevic


-----------------------------------------
The information contained in this message is intended only for the personal and 
confidential use of the recipient(s) named above. If the reader of this message 
is not the intended recipient or an agent responsible for delivering it to the 
intended recipient, you are hereby notified that you have received this 
document in error and that any review, dissemination, distribution, or copying 
of this message is strictly prohibited. If you have received this communication 
in error, please notify us immediately, and delete the original message.

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