I can tell you my understanding - right or wrong!
Operating systems keep lots of I/O queues . . at least these:
- a queue for each HBA.
- a queue for each LUN.
By spreading your db across multiple luns and hba's, you can get more
parallelism - assuming your application can make use of the parallelism.
This is why TSM also should have it's database defined using multiple
volumes. So you get multiple db vols across multiple lun's across multiple
HBA's.
Will it help? It all depeneds . . .
The best thing I have ever read about this topic is the following. It
doesn't specifically address your question, but it goes into a deep dive
about queuing.
http://virtualgeek.typepad.com/virtual_geek/2009/06/vmware-io-queues-micro-bursting-and-multipathing.html
Rick
Henrik Vahlstedt
<shwl AT STATOIL DOT COM
> To
Sent by: "ADSM: ADSM-L AT VM.MARIST DOT EDU
Dist Stor cc
Manager"
<[email protected] Subject
.EDU> disk question
04/21/2010 11:11
AM
Please respond to
"ADSM: Dist Stor
Manager"
<[email protected]
.EDU>
Hi,
Are there any performance benefits in creating two or more LUN´s in the
same RAID group for one RHEL 5, TSM 5.5 database, instead of only creating
one LUN. If yes, why?
Example: IBM Midrange System Storage Implementation and Best Practices
Guide sg246363, see chapter 7.4/TSM database.
//Henrik
-------------------------------------------------------------------
The information contained in this message may be CONFIDENTIAL and is
intended for the addressee only. Any unauthorised use, dissemination of the
information or copying of this message is prohibited. If you are not the
addressee, please notify the sender immediately by return e-mail and delete
this message.
Thank you
|