Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: ORACLE on Linux - IO bottleneck

Re: ORACLE on Linux - IO bottleneck

From: Wyvern <edersm_at_wanadoo.es>
Date: 9 Feb 2006 07:52:30 -0800
Message-ID: <1139500350.096649.226750@o13g2000cwo.googlegroups.com>

> Hi Mladen,
> actually the I/O scheduler(s) works even if the I/O request are not
> simultaneous.
> typical is the merging of I/O for sequential reads or writes.
>
> The advantages can be tremendous.
>
> Here is an example (same hardware, same device):
>

......

> The second dd has its I/O merged.
>
> Consider this controversial but the I/O scheduler eliminate the need to
> have oracle block size "aligned" to file system block size (even without
> direct I/O).
>

It´s interesting what you say and the example shown. This morning, looking about diferent IO parameters at device level with "blockdev" command, I´ve seen, we´ve a blocksize of 1K in the device and, as I told you, 120 readahead.

As you say in your previous post, without an IO scheduler, we may have a problem if we have an oracle 8K block size and a device (sdc, sdd, etc ...) 1K blocksize. Is it right?. Before reading your post, I thought that using direct IO let us not worrying about device´s blocksize but .... now I´m in doubt...

Could it be another point to consider in this case?

What do you recommend: "aligning" both block sizes (we have unique oracle block size - 8K) or using an IO scheduler (IO elevator as you said previously in kernel 2.4) ?? Received on Thu Feb 09 2006 - 09:52:30 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US