[K12OSN] Dell PE 2500 / Performance

John Baillie (jsbaillie@cfl.rr.com)
22 Mar 2002 19:17:26 -0500 


--=-FKMDF+zsNKswCVzoA63o
Content-Type: text/plain
Content-Transfer-Encoding: 7bit

Doug, 

Matt Domsch is the senior software engineer for Dell Linux and has
posted good Dell info at http://domsch.com/linux

Prior to going live with K12LTSP we had K12 Ver 1.x installed on our PE
2500 it screamed. When we loaded Ver 2.0 and there was a noticeable drop
in performance. The percent user processes spikes and hangs at 85% to
95% at log-on with 30 concurrent users and stays there for about 5
minutes. 

Using bonnie++ to test disk IO we had less than stunning results. I
don't have the specs with me, but will post them if there is any
interest. 

I know time is at a premium for most of us but I believe it would be
useful to do some bench marks using an agreed upon set of tools and
conditions. 

On k12ltsp there is a note saying no system is going to perform well
with 30 concurrent log-ons. Maybe we'll have to accept that. As a work
around we have the students log on a row at a time. 

Regards,
John

On Fri, 2002-03-22 at 16:13, Doug Simpson wrote: 

> We are running a Dell PE 2400 server here, and RH6.2.  No problems that
> I can tell. . .
> 
> DS
> 
> > We are using a Dell PE 2500 we got a very good deal on it. But we have
> > noticed a performance bottle neck with the perc 3/di scsi raid
> > controller and the latest kernel. If you go with Dell you might want
> to
> > consider choosing another controller. If I was going to do it all over
> > again I'd build my own. 
> > 
> > Regards,
> > John Baillie
snip--------------------------------------------------------


--=-FKMDF+zsNKswCVzoA63o
Content-Type: text/html; charset=utf-8




  
  


Doug, 





Matt Domsch is the senior software engineer for Dell Linux and has posted good Dell info at http://domsch.com/linux





Prior to going live with K12LTSP we had K12 Ver 1.x installed on our PE 2500 it screamed. When we loaded Ver 2.0 and there was a noticeable drop in performance. The percent user processes spikes and hangs at 85% to 95% at log-on with 30 concurrent users and stays there for about 5 minutes. 





Using bonnie++ to test disk IO we had less than stunning results. I don't have the specs with me, but will post them if there is any interest. 





I know time is at a premium for most of us but I believe it would be useful to do some bench marks using an agreed upon set of tools and conditions. 





On k12ltsp there is a note saying no system is going to perform well with 30 concurrent log-ons. Maybe we'll have to accept that. As a work around we have the students log on a row at a time. 





Regards,


John





On Fri, 2002-03-22 at 16:13, Doug Simpson wrote: 

> We are running a Dell PE 2400 server here, and RH6.2.  No problems that
> I can tell. . .
> 
> DS
> 
> > We are using a Dell PE 2500 we got a very good deal on it. But we have
> > noticed a performance bottle neck with the perc 3/di scsi raid
> > controller and the latest kernel. If you go with Dell you might want
> to
> > consider choosing another controller. If I was going to do it all over
> > again I'd build my own. 
> > 
> > Regards,
> > John Baillie
snip--------------------------------------------------------





--=-FKMDF+zsNKswCVzoA63o--



_______________________________________________
K12OSN mailing list
K12OSN@redhat.com
https://listman.redhat.com/mailman/listinfo/k12osn
For more info see