On Sunday July 30, [EMAIL PROTECTED] wrote:
> Hello
>
> I gave 2.4.0-test5 a try but am very disapointed with the results:
>
> -------Sequential Output-------- ---Sequential Input-- --Random--
> -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks---
> Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU
> 2.2.15 384 5231 95.2 16051 27.3 7557 23.9 5764 95.0 18547 18.3 162.2 2.6
> 2.4.0-test5 384 1568 27.2 1700 1.9 1084 3.8 2273 41.2 5326 10.5 162.7 4.0
>
> This is on a 2xPII-350 with 128MB and 5 disks (1 IDE, 4 old IBM SCSI disks)
> on a Software Raid5.
>
> Do other people get similar results or am I doing something wrong?
The difference is steeper than I would expect ... but maybe my
expectations are wrong:-)
the raid5 code in 2.4 is quite different from the patches for 2.2,
partly because of the need to cope with SMP issues. The code
is currently over cautious and blocks where maybe it doesn't need to.
I hope to fix this up, but it all takes time.
NeilBrown
