Andrew Edunov wrote:
> Hi all!
>
> I'm working on IBM x365 + SuSE 9 kernel-bigsmp-2.6.5-7.282 + Oracle
> 10.2.0.2.
> I have a huge workloads on this server and time after time I look at sar
> statistic on this server:
>
>             CPU   %user   %nice  %system  %iowait   %idle
> 15:00:02       all   64.76    0.00   35.23    0.00    0.00
> 15:10:01       all   64.65    0.00   35.31    0.04    0.01
> 15:20:02       all   64.63    0.00   35.35    0.01    0.00
> 15:30:06       all   67.83    0.00   32.17    0.00    0.00
> 15:40:02       all   63.03    0.00   36.97    0.00    0.00
> 15:50:05       all   62.22    0.00   37.78    0.00    0.00
> 16:00:04       all   63.89    0.00   36.11    0.00    0.00
> 16:10:01       all   62.47    0.00   37.15    0.04    0.33
> 16:20:03       all   58.30    0.00   34.54    0.72    6.43
> 16:30:01       all   64.53    0.00   35.39    0.06    0.01
> 16:40:03       all   63.49    0.00   35.93    0.18    0.40
> 16:50:01       all   63.37    0.00   36.62    0.00    0.01
> 17:00:02       all   63.44    0.00   36.55    0.00    0.00
> 17:10:02       all   63.06    0.00   36.94    0.00    0.00
> 17:20:02       all   61.90    0.00   38.07    0.03    0.00
> 17:30:01       all   64.21    0.00   33.47    0.39    1.93
> 17:40:01       all   59.07    0.00   34.90    0.47    5.57
> 17:50:01       all   62.63    0.00   37.27    0.09    0.01
> 18:00:09       all   64.41    0.00   35.59    0.00    0.00
> 18:10:01       all   63.59    0.00   36.20    0.07    0.15
> 18:20:01       all   62.12    0.00   35.65    0.18    2.06
> 18:30:01       all   58.35    0.00   31.71    2.05    7.90
> 18:40:04       all   61.33    0.00   33.66    0.45    4.55
> 18:50:01       all   62.04    0.00   35.60    0.46    1.91
>
> And for me this looks strange. Utilization on user level processes takes
> only 60% and kernel level utilization takes 40%. Is this normal? Or may be
> this is some kernel bug or something?
>
> Best regards,
>  Andrew Edunov.
>
>  

I've seen this behaviour before on UNIX boxes.
It can happen when the CPU('s) are so heavily saturated that the OS is
spending
a lot of time doing scheduling and process management instead of letting
the user
processes do the real work.

There is however also the possibility that your user processes initiates
some kernel
level operations.

Have you taken a look at top while the load is as described above?


--
Best regards

Martin Berg

email: martin@(protected)
web: www.berg-consult.com


--
To unsubscribe, email: suse-oracle-unsubscribe@(protected)
For additional commands, email: suse-oracle-help@(protected)
Please see http://www.suse.com/oracle/ before posting