RE: OT: high WIO on Linux

From: Kevin Closson <kevinc_at_polyserve.com>
Date: Wed, 15 Mar 2006 12:34:10 -0800
Message-ID: <5D2570CAFC98974F9B6A759D1C74BAD06FE4C0@ex2.ms.polyserve.com>

 >>>I've always wondered that. How can you determine that? Is 

>>>the iostat average IO size per device any good for this ?
>>>
>>>Sometimes even if it does chew it up in 64k requests, I find
>>>it's better to have it large, so that you can queue up a
>>>large number of IOs in a serial read process.
>>>

iostat -x will help you resolve that. Some of the many SANs we have inhouse here also report out what the average transfer size is...and if I'm doing, say, a CCF or LWS, I should be hitting the array at 1MB, but if I'm on something like RH3 (puke all over please so I don't have to) it'll be more like 32KB or 64KB "chunklets"...

The 2.6 kernel is not so messed up this way. You are right too, that even if it gets chopped down into smaller requests, it is still better to have Oracle through out the larger requests.

--
http://www.freelists.org/webpage/oracle-l

Received on Wed Mar 15 2006 - 14:34:10 CST