Date: Thu, 10 Aug 2006 14:32:18 +1000 (EST) From: Ian Smith <smithi@nimnet.asn.au> To: questions@freebsd.org Subject: Top not showing cpu usage even remotely accurately Message-ID: <Pine.BSF.3.96.1060810135055.16843A-100000@gaia.nimnet.asn.au>
next in thread | raw e-mail | index | archive | help
In the past (FreeBSD 2.2, 3.3, 4.5 through 4.10) if I noticed unexpected
high cpu use, say on xload or ascpu, I'd just run top to determine what
the pig was, and kill it if something (usually netscape :) was wedged.
But since running 5.x (5.5-STABLE since 1st Aug) top can show 0.0% idle
but the cpu usages shown don't add up to much of a fraction of 100%.
For a typical illustration: 'make index' has been running for hours, and
here's a shot of 'nice top', o)rdered by cpu, showing S)ystem procs:
last pid: 62397; load averages: 2.06, 2.09, 2.13 up 3+03:57:39 13:11:27
154 processes: 6 running, 125 sleeping, 23 waiting
CPU states: 77.3% user, 0.0% nice, 22.7% system, 0.0% interrupt, 0.0% idle
Mem: 79M Active, 24M Inact, 37M Wired, 5640K Cache, 25M Buf, 3388K Free
Swap: 384M Total, 133M Used, 251M Free, 34% Inuse
PID USERNAME PRI NICE SIZE RES STATE TIME WCPU CPU COMMAND
62380 root 122 0 788K 648K RUN 0:01 19.48% 1.86% make
736 smithi 97 0 27000K 3184K select 89:56 1.07% 1.07% kdeinit
762 smithi 96 0 29708K 7428K select 14:01 0.93% 0.93% kdeinit
699 smithi 96 0 52320K 10288K select 105:02 0.88% 0.88% Xorg
62394 root 125 0 580K 440K RUN 0:00 7.00% 0.34% make
754 smithi 96 0 30972K 6308K select 57:22 0.15% 0.15% kdeinit
770 smithi 8 0 2524K 660K nanslp 20:51 0.05% 0.05% ascpu
11 root 171 52 0K 8K RUN 60.4H 0.00% 0.00% idle
760 smithi 20 0 10636K 1404K kserel 187:08 0.00% 0.00% xmms
27 root -28 -147 0K 8K RUN 27:50 0.00% 0.00% swi5: clock sio
757 smithi 96 0 28784K 4220K select 26:03 0.00% 0.00% kdeinit
767 smithi 96 0 29828K 5704K select 24:01 0.00% 0.00% kdeinit
28643 smithi 20 0 71016K 30816K kserel 23:01 0.00% 0.00% mozilla-bin
769 smithi 8 0 2300K 612K nanslp 21:23 0.00% 0.00% asapm
639 root 96 0 1244K 364K select 17:02 0.00% 0.00% moused
761 smithi 96 0 29468K 5132K select 16:32 0.00% 0.00% kdeinit
763 smithi 96 0 36568K 5940K select 13:06 0.00% 0.00% kdeinit
619 mysql 20 0 49800K 1528K kserel 12:27 0.00% 0.00% mysqld
752 smithi 96 0 29644K 4976K select 10:59 0.00% 0.00% kdeinit
759 smithi 96 0 28720K 4548K select 9:34 0.00% 0.00% kdeinit
15822 smithi 8 0 1488K 412K nanslp 4:30 0.00% 0.00% ephem
750 smithi 96 0 27304K 5316K select 3:03 0.00% 0.00% kdeinit
4 root -8 0 0K 8K - 1:50 0.00% 0.00% g_down
23 root -60 -179 0K 8K WAIT 1:31 0.00% 0.00% irq12: psm0
3 root -8 0 0K 8K - 1:30 0.00% 0.00% g_up
43 root 20 0 0K 8K syncer 1:28 0.00% 0.00% syncer
49 root 12 0 0K 8K - 1:16 0.00% 0.00% schedcpu
30 root -16 0 0K 8K - 1:06 0.00% 0.00% yarrow
39 root -16 0 0K 8K psleep 0:56 0.00% 0.00% pagedaemon
16 root -80 -199 0K 8K WAIT 0:48 0.00% 0.00% irq5: sbc0
2 root -8 0 0K 8K - 0:48 0.00% 0.00% g_event
41 root 171 52 0K 8K RUN 0:46 0.00% 0.00% pagezero
29 root -44 -163 0K 8K WAIT 0:26 0.00% 0.00% swi1: net
545 root 96 0 9764K 492K select 0:21 0.00% 0.00% httpd
25 root -64 -183 0K 8K WAIT 0:21 0.00% 0.00% irq14: ata0
13 root -60 -179 0K 8K WAIT 0:18 0.00% 0.00% irq1: atkbd0
490 root 96 0 3464K 744K select 0:17 0.00% 0.00% sendmail
[etc]
'ps auxww' cpu percentages reveal little more, except that make index is
running 'make -j2 ..' hence the ~2.0 load average.
Ignore the high swap use; most of it is numerous quiescent kwrite, httpd
and mozilla sessions pushed out to swap on this lil' 160MB laptop;
'systat -vm' shows it's not actually doing any paging during this time.
Any ideas why top hasn't much of a clue about what's consuming cpu?
Cheers, Ian
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.3.96.1060810135055.16843A-100000>