Re: memory and page faults (continued - again)


Server in question.
12:root:/ > ps -k
PID TTY TIME CMD
0 - 46:29 swapper
516 - 41004:57 kproc
774 - 40133:39 kproc
1032 - 40094:22 kproc
1290 - 39980:45 kproc
1548 - 39797:17 kproc
1806 - 39755:31 kproc
2064 - 688:51 kproc
2322 - 0:25 kproc
2580 - 64:10 kproc
2838 - 0:21 kproc
4938 - 0:00 kproc
5422 - 0:00 kproc
6494 - 0:01 kproc
8006 - 0:02 kproc
8520 - 191:47 kproc
12402 - 0:04 kproc
14048 - 191:25 kproc
19610 - 191:16 kproc
19870 - 191:22 kproc
20128 - 191:05 kproc
20382 - 191:11 kproc
20640 - 190:39 kproc
20898 - 190:47 kproc
21156 - 191:10 kproc
21414 - 191:00 kproc
202120 - 0:00 kproc
12:root:/ > ps -k | wc -l
28

Compared to AIX 5.1 server output:
14:ROOT > ps -k
PID TTY TIME CMD
0 - 24:11 swapper
516 - 55498:36 wait
774 - 51271:58 wait
1032 - 50990:30 wait
1290 - 50756:08 wait
1548 - 0:09 reaper
1806 - 409:20 lrud
2064 - 0:30 xmgc
2322 - 0:28 netm
2580 - 9:54 gil
2838 - 0:00 wlmsched
4152 - 0:00 lvmbb
4932 - 1:09 rt-fcparray
5468 - 54:40 aioserver
6712 - 0:00 dog
7286 - 0:01 rtcmd
8010 - 0:00 kbiod
8522 - 0:06 j2pg
11354 - 5:29 nfsd
11620 - 5:29 rpc.lockd
15742 - 0:06 PM
21692 - 54:38 aioserver
22260 - 54:49 aioserver
24272 - 54:47 aioserver
24774 - 54:55 aioserver
25292 - 54:51 aioserver
26062 - 54:53 aioserver
26316 - 54:38 aioserver
26574 - 54:41 aioserver
26832 - 54:47 aioserver
45482 - 0:00 jfsz
You have mail in /usr/spool/mail/root
14:ROOT > ps -k | wc -l
32
14:ROOT > ps -k | grep aio | wc -l
10

Holger.VanKoll@xxxxxxxxxxxx 02/03/06 7:17 am >>>

ps -k shows all kprocs, aio or not

On that server I can see I have 32 and 10 of those are aioservers so is that what I'm looking for?
I dont understand that question. Just look at the aio-servers (their cpu-load). Maybe provide ps -k output to get more help.

Just curious on the 'lots of sequential reads comment' was that gleaned from the filemon output I provided earlier?
Yes.
You had hundreds of reads any only 2 read-sequences. That means all those hundres of reads where read with 2 times positioning the read/write head.
So far I havent seen such highly sequential read-patterns on databases in production environments


From: IBM AIX Discussion List [mailto:aix-l@xxxxxxxxxxxxx] On Behalf Of Mark Schlechte
Sent: Thursday, February 02, 2006 6:37 PM
To: aix-l@xxxxxxxxxxxxx
Subject: Re: memory and page faults (continued - again)


I think I need a little clarification please.
ps -k gives me 28 kproc processes (aix 433 server).
ps -k on an AIX 5L server gives me more info on the actual kernel process.
On that server I can see I have 32 and 10 of those are aioservers so is that what I'm looking for?
Just curious on the 'lots of sequential reads comment' was that gleaned from the filemon output I provided earlier?


Holger.VanKoll@xxxxxxxxxxxx 02/02/06 6:56 am >>>

you need to run cfgmgr (aix5) to make the changes of aioservers take effect
on 4.x you have to reboot if I remember correctly



DISCLAIMER: The information transmitted is intended only
for the addressee and may contain confidential,
proprietary and/or privileged material. Any
unauthorized review, distribution or other use
of or the taking of any action in reliance upon
this information is prohibited. If you received
this in error, please contact the sender and
delete or destroy this message and any copies.

As of January 20, any e-mails addressed
to "cityregina.com" will not be delivered.