 | | | increased I/O tps | increased I/O tps 2004-03-01 - By Christopher McCrory
Back
Hello...
On Mon, 2004-03-01 at 14:58, Stephen Smoogen wrote:
> On Mon, 2004-03-01 at 15:46, Christopher McCrory wrote:
> > Hello...
> >
> > I 'm wondering if anyone else is seeing this or knows the reason for the
> > following.
> >
> > two identical servers, Intel, SMP, Adaptec, software raid, U160 scsi,
> > same workload. One is RH7.3 (all updates) , the other is RHEL3 (all
> > updates).
> >
>
> What kind of adaptec card?
>
Adaptec AIC-7899P U160/m
Dell 2650s
(No PERC )
> > according to iostat, the tps is roughly double on RHEL3. BLk_read/write
> > is the same. load and cpu usage is roughly double.
> >
> > It seems as if all the scsi reads/writes are half the size they were
> > resulting in twice the transactions. yes? no? maybe? other?
> >
> > I 've tried reverting to 2.4.21-4.0.2.ELsmp, elvtune, disabling HT, and
> > mem=4096M (down from 6G), all to no avail. ? ? ?
> >
> > Or should I open a support ticket?
>
> What exactly is the problem :)?
I was trying to show the general idea without revealing too much
internal info :)
The problem is higher disk I/O probably causing higher load and CPU
usage with RHEL3 vs. RH7.3
> Reference numbers to compare to might be
> helpful... but if you have support, by all means open a ticket. The main
> thing is to be able to ask the question in a way that cuts down the
> support time :).
When possible I prefer a public forum for support as I can solve most
issues by looking at list archives. Maybe this is something simple I
missed and the answer will help someone else down the road.
--
Christopher McCrory
"The guy that keeps the servers running "
chrismcc@(protected)
http://www.pricegrabber.com
Let 's face it, there 's no Hollow Earth, no robots, and
no 'mute rays. ' And even if there were, waxed paper is
no defense. I tried it. Only tinfoil works.
--
Taroon-list mailing list
Taroon-list@(protected)
http://www.redhat.com/mailman/listinfo/taroon-list
|
|
|