LVS
lvs-users
Google
 
Web LinuxVirtualServer.org

RE: Question About SMP

To: 'Peter Mueller' <pmueller@xxxxxxxxxxxx>, <lvs-users@xxxxxxxxxxxxxxxxxxxxxx>
Subject: RE: Question About SMP
From: Dusten Splan <Dusten@xxxxxxxxxxxxxxxxxx>
Date: Wed, 14 May 2003 15:16:02 -0700
I thought that this was solved with the 2.4 kernel, because it was meant to
be multithreaded.  I'm running 2.4.20 with the newest LVS. Maybe I need to
run 2.5 but this is a production machine. Has anyone tried LVS on 2.5? 

I have tried the irqbalance and that doesn't seam to be helping any, Oh and
does anyone know of a web page for that patch.  I guess I just might get a
3GHz box and forget the hole smp thing.  Here's a sample of what mpstat has
to say about the hole thing.

[root@www99 root]# mpstat -P ALL 1 10; 
Linux 2.4.20 (www99)    05/14/2003

03:10:05 PM  CPU   %user   %nice %system   %idle    intr/s
03:10:06 PM  all    0.00    0.00   12.00   88.00  19841.00
03:10:06 PM    0    0.00    0.00   23.00   77.00  19748.00
03:10:06 PM    1    0.00    0.00    1.00   99.00    100.00

03:10:06 PM  CPU   %user   %nice %system   %idle    intr/s
03:10:07 PM  all    0.00    0.00   10.00   90.00  20532.00
03:10:07 PM    0    0.00    0.00   19.00   81.00  20428.00
03:10:07 PM    1    0.00    0.00    1.00   99.00    100.00

03:10:07 PM  CPU   %user   %nice %system   %idle    intr/s
03:10:08 PM  all    0.00    0.00    9.50   90.50  19738.00
03:10:08 PM    0    0.00    0.00   18.00   82.00  19640.00
03:10:08 PM    1    0.00    0.00    1.00   99.00    100.00

03:10:08 PM  CPU   %user   %nice %system   %idle    intr/s
03:10:09 PM  all    0.00    0.00   10.50   89.50  20149.00
03:10:09 PM    0    0.00    0.00   20.00   80.00  20043.00
03:10:09 PM    1    0.00    0.00    1.00   99.00    100.00

-----Original Message-----
From: Peter Mueller [mailto:pmueller@xxxxxxxxxxxx]
Sent: Wednesday, May 14, 2003 2:57 PM
To: 'LinuxVirtualServer.org users mailing list.';
'dusten@xxxxxxxxxxxxxxxxxx'
Subject: RE: Question About SMP


Hello Dusten,

> When doing more traffic the load on cpu0 increases and 
> nothing is happing on
> cpu1. 
> 
> My question is why am I not seeing this processor usage 
> distributed over
> both processors.  I know that on a Sun box the network card 
> is stuck to a
> single processor and will not use the other processors.  

Uhm... what version of LVS are you using?

I remember quite a few discussions indicating SMP was not `efficient` with
LVS due to spinlocks & such.  The end result was that it was better to
compile without SMP.  I'm not sure why you aren't seeing utilization on both
CPUs.
<Prev in Thread] Current Thread [Next in Thread>