From: ldg@ulysium.net
Date: Wed Jun 11 2003 - 16:04:08 EDT
Hello all again,
yesterday I posted about this small issue of system usage:
> USER PID %CPU %MEM VSZ RSS TTY S STARTED TIME
> COMMAND
> root 0 12.5 2.9 1.37G 37M ?? R < 11:58:09 49:16.12
> [kernel idle]
>
>
> load average: 0.14, 0.10, 0.10
>
> The load average stay a lot around 0.10 and 0.20 even when most tasks are
> shut down and there's no load on the system.
I have gone through searching the list archives and google but although I
found a little info from older posts, I didn't find anything that helped me
track down the culprit yet.
I did try to use ps to look up the threads but it hasn't helped because it's
not giving the cpu usage per thread and more importantly not always saying
what each thread is used by.
Here's an example of ps that I did:
ps axml -p 0
UID PID PPID CP PRI NI VSZ RSS WCHAN S TTY
TIME COMMAND
0 0 0 131 38 -6 1.37G 37M * R < ??
03:11:15 [kernel idle]
0 63 19 - R N
0:00.00
0 44 0 - R
0:00.00
0 38 -6 malloc_ S <
2:25.09
0 38 -6 180a0d8 U <
0:00.04
0 32 -12 ubc_pur I <
0:00.08
0 37 -7 Tswpin I <
0:00.00
0 37 -7 180a418 U <
0:00.00
0 37 -7 180a498 U <
0:00.00
0 38 -6 861908 I <
0:01.48
0 38 -6 863118 S <
0:00.01
0 38 -6 863158 I <
0:00.00
0 63 19 - R N
0:00.00
0 37 -7 a40280 U <
0:00.06
0 32 -12 9f1d88 U <
0:00.00
0 32 -12 9f1540 U <
0:00.00
0 32 -12 63f270 U <
0:00.00
0 32 -12 62f2c0 U <
0:00.07
0 32 -12 62f500 U <
0:00.00
4 32 -12 camwork S <
7:58.63
0 32 -12 camwork I <
0:00.00
0 32 -12 a01da0 U <
0:00.00
0 38 -6 ss_tmo S <
0:00.05
2 31 -13 isp_rq S <
3:46.75
0 38 -6 isp_abo I <
0:00.00
0 38 -6 isp_fm I <
0:00.00
0 38 -6 ss_tmo S <
0:00.08
0 38 -6 863ca0 U <
0:02.47
3 31 -13 isp_rq S <
3:45.90
0 38 -6 isp_abo I <
0:00.00
0 38 -6 isp_fm I <
0:00.01
0 38 -6 ss_tmo S <
0:00.12
0 38 -6 85eb50 U <
0:00.59
0 32 -12 6b7a6c U <
0:00.00
0 38 -6 8e337aa8 U <
0:00.06
0 41 -3 a40448 U <
0:00.00
0 33 -11 netisr S <
0:03.91
0 33 -11 netisr S <
0:03.87
0 33 -11 netisr S <
0:03.83
0 33 -11 tcpfstm S <
0:00.25
0 33 -11 tcpsltm S <
0:00.25
0 33 -11 574879f0 S <
0:00.88
0 38 -6 netrain I <
0:00.00
0 38 -6 netrain S <
0:00.02
0 38 -6 netrain I <
0:00.00
0 33 -11 icmp6_s S <
0:00.20
0 33 -11 icmp6_f S <
0:00.19
0 38 -6 ranbat I <
0:00.13
0 38 -6 8638a0 U <
0:00.03
1 38 -6 8de73e10 U <
1:52.69
0 38 -6 8de73db0 U <
0:00.00
0 38 -6 8de73d50 U <
0:00.01
0 38 -6 8de73cf0 U <
0:00.00
0 38 -6 8585e0 U <
0:00.00
0 38 -6 8de73c90 U <
0:00.00
0 38 -6 8de73c30 U <
0:00.01
0 38 -6 8de73bd0 U <
0:24.46
0 38 -6 8de73bd0 U <
0:24.91
0 38 -6 8de73b70 U <
01:17:07
118 38 -6 1b932c0 U <
01:16:16
0 38 -6 - U <
0:00.31
0 38 -6 8cc8c0 U <
0:00.00
0 0 -44 8cd250 U <
0:00.00
0 42 -2 26f290 U <
0:00.01
0 38 -6 85a3c0 U <
0:15.73
1 38 -6 9ef868 U <
8:15.07
2 38 -6 9ef868 U <
8:14.63
0 38 -6 86051c U <
0:00.00
0 38 -6 2be08180 U <
0:00.00
0 38 -6 ea5ad40 U <
0:00.00
0 38 -6 ea5ad00 U <
0:00.00
This isn't very helpful so far, unless I'm not reading something in this
that is relevant (quite possible).
One thing of importance that came up and didn't mention in my earlier post,
is that I am not using nfs and I do have it in the kernel. Same thing with
netrain, I'm not using it but it's there, although I doubt netrain would
consume cpu like this anyway.
Any suggestions as to what else to look at?
-- Didier Godefroy mailto:dg@ulysium.net
This archive was generated by hypermail 2.1.7 : Sat Apr 12 2008 - 10:49:22 EDT