Followup question: Question concerning high run queue / low CPU Utilization

From: Brewer, Edward (NIH/OD) (BREWERE@OD.NIH.GOV)
Date: Tue Mar 25 2003 - 14:54:18 EST


Follow-up question:

I do notice that the value for NLWP using ps -efL is around 650 and 950 for
two process. I think this sounds like an unseemly large number, does this
sound right?

TIA,
Lee

> To all,
>
>
> I notice very large values for sar -q on one of our Solaris 8 machines.
>
> Here is some output
>
>
> 00:00:00 runq-sz %runocc swpq-sz %swpocc
> 08:26:00 97.0 1
> 08:28:00 638.3 2
> 08:30:00 23.0 1
> 08:32:00 633.2 3
> 08:34:00 298.2 4
> 08:36:00 484.2 4
> 08:38:00 509.2 3
> 08:40:00 1015.0 1
> 08:42:00 493.9 12
> 08:44:00 444.9 7
> 08:46:00 490.4 8
> 08:48:00 534.2 3
> 08:50:00 509.0 10
> 08:52:00 575.2 5
> 08:54:00 333.7 9
> 08:56:00 453.8 11
> 08:58:00 402.9 13
> 09:00:00 308.6 7
> 09:02:00 484.7 12
> 09:04:00 593.0 6
> 09:06:00 335.7 6
> 09:08:00 535.5 18
>
> With this for CPU Usage
>
> 08:26:00 0 1 0 98
> 08:28:00 1 3 0 95
> 08:30:00 0 2 0 98
> 08:32:00 1 4 0 95
> 08:34:00 1 4 0 95
> 08:36:00 1 3 0 96
> 08:38:00 1 5 0 94
> 08:40:00 1 4 0 95
> 08:42:00 2 9 0 89
> 08:44:00 1 6 0 93
> 08:46:00 1 7 0 92
> 08:48:00 1 3 0 97
> 08:50:00 2 8 0 90
> 08:52:00 2 6 0 92
>
> We are running Oracle 9ias release 1 and 2 (OC4J). I believe that we have
> the lastest patches. The disk values from sar appear good ( I can send
> some if you really want it) and free memory never really drops that low.
>
>
> Am I seeing large numbers due to threading????
>
> TIA,
> Lee Brewer
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers



This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 23:26:03 EDT