Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.tools -> kproc

kproc

From: Nicholas Buckley <nbuck_at_globalnet.co.uk>
Date: 1999/04/28
Message-ID: <7g7mo3$l6$1@gxsn.com>#1/1

Hi everyone,

Question [ sorry, 4 questions :-) ] concerning one of four nodes in a Bull EPC 400 cluster running HACMP under AIX 4.3.1.0.

Any insight would be warmly welcomed.

The AIX Survival Guide says:

"Usually you will find one of these kprocs ( PID 516 ) which eats up a great deal of CPU. This is the wait process. Any time the operating system has nothing to do, the wait process is run. The architecture of AIX requires that at least one process is always running."

Ok, but ps -efk | grep kproc returns:

    root   516     0 120   10 Apr      - 21062:07 kproc
    root   774     0 120   10 Apr      - 21064:48 kproc
    root  1032     0 120   10 Apr      - 21525:00 kproc
    root  1290     0 120   10 Apr      - 21630:50 kproc
    root  1548     0   0   10 Apr      - 181:03 kproc
    root  1806     0   0   10 Apr      -  0:44 kproc
    root  2064     0   0   10 Apr      - 100:06 kproc
    root  2988     0   0   10 Apr      -  0:00 kproc
    root  3412     1   0   10 Apr      -  0:00 kproc
    root  4160     1   0   10 Apr      -  0:00 kproc
    root  5990     1   0   10 Apr      -  0:00 kproc
    root 15182     1   0   10 Apr      -  0:00 kproc
    root 19156     0   0   12 Apr      -  0:00 kproc
    root 52426     0   0   12 Apr      -  0:02 kproc

ps aux | grep kproc
USER PID %CPU %MEM SZ RSS TTY STAT STIME TIME COMMAND

root      1290 21.3  0.0    4    8      - A      10 Apr 21653:53 kproc
root      1032 21.2  0.0    4    8      - A      10 Apr 21547:55 kproc
root       774 20.7  0.0    4    8      - A      10 Apr 21087:26 kproc
root       516 20.7  0.0    4    8      - A      10 Apr 21083:29 kproc
root      1548  0.2  0.0    8    8      - A      10 Apr 181:03 kproc
root      2064  0.1  0.0   60   56      - A      10 Apr 100:13 kproc
root      1806  0.0  0.0   12   16      - A      10 Apr  0:44 kproc
root     52426  0.0  0.0   12   16      - A      12 Apr  0:02 kproc
root      2988  0.0  0.0   12    8      - A      10 Apr  0:00 kproc
root      3412  0.0  0.0   12   16      - A      10 Apr  0:00 kproc
root      5990  0.0  0.0   12    8      - A      10 Apr  0:00 kproc
root      4160  0.0  0.0   12    8      - A      10 Apr  0:00 kproc
root     15182  0.0  0.0   12    8      - A      10 Apr  0:00 kproc
root     19156  0.0  0.0   12    8      - A      12 Apr  0:00 kproc

Question 1)
Why are there so many kprocs?

Question 2)
Is this normal ?

Question 3)
What does this tell me about the way the system is performing?

Question 4)
The first four entries under ps aux are using up over 80% of CPU time. Can I do anything about this ( especially as the other 10 are using almost no CPU at all )

Regards,

Nick Buckley
AIX Administrator,
NCM,
Cardiff,
Wales,
United Kingdom
e-mail: nicholas.buckley_at_ncmgroup.com
(Please note: all newsgroup offerings are made in a personal capacity and in no way are the responsibility of my employer) Received on Wed Apr 28 1999 - 00:00:00 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US