From: "Metathronius Galabant" <m.galabant@googlemail.com>
To: "Andrew Morton" <akpm@osdl.org>
Cc: linux-kernel@vger.kernel.org
Subject: Re: top displaying 50% si time and 50% idle on idle machine
Date: Fri, 8 Sep 2006 23:03:31 +0200 [thread overview]
Message-ID: <1b270aae0609081403u11b76ae9v72ad933475a2319f@mail.gmail.com> (raw)
In-Reply-To: <20060907175323.57a5c6b0.akpm@osdl.org>
>> Kernel 2.6.17.11 with tg3 network driver, NAPI enabled (Distro CentOS 4.4).
>> top shows strangely 50% idle even if the machine is _completely_ idle.
>>
>> top - 01:04:30 up 4 days, 12:05, 2 users, load average: 0.00, 0.00, 0.00
>> Tasks: 34 total, 2 running, 32 sleeping, 0 stopped, 0 zombie
>> Cpu(s): 0.0% us, 0.0% sy, 0.0% ni, 50.0% id, 0.0% wa, 0.0% hi, 50.0%si
> Do `ps aux', look for a process stuck in D state.
The issue that startled me: there is _NO_ process in a D state!
BTW what means si? (interrupt service time? google didn't find anything)
ps auxwm output following:
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
root 1 0.0 0.0 1692 572 ? - Sep04 0:00 init [3]
root - 0.0 - - - - Ss Sep04 0:00 -
root 2 0.0 0.0 0 0 ? - Sep04 0:00 [ksoftirqd/0]
root - 0.0 - - - - SN Sep04 0:00 -
root 3 0.0 0.0 0 0 ? - Sep04 0:00 [events/0]
root - 0.0 - - - - S< Sep04 0:00 -
root 4 0.0 0.0 0 0 ? - Sep04 0:00 [khelper]
root - 0.0 - - - - S< Sep04 0:00 -
root 5 0.0 0.0 0 0 ? - Sep04 0:00 [kthread]
root - 0.0 - - - - S< Sep04 0:00 -
root 7 0.0 0.0 0 0 ? - Sep04 0:00 [kblockd/0]
root - 0.0 - - - - S< Sep04 0:00 -
root 8 0.0 0.0 0 0 ? - Sep04 0:00 [kseriod]
root - 0.0 - - - - S< Sep04 0:00 -
root 70 0.0 0.0 0 0 ? - Sep04 0:00 [pdflush]
root - 0.0 - - - - S Sep04 0:00 -
root 71 0.0 0.0 0 0 ? - Sep04 0:00 [pdflush]
root - 0.0 - - - - S Sep04 0:00 -
root 72 0.0 0.0 0 0 ? - Sep04 0:00 [kswapd0]
root - 0.0 - - - - S Sep04 0:00 -
root 73 0.0 0.0 0 0 ? - Sep04 0:00 [aio/0]
root - 0.0 - - - - S< Sep04 0:00 -
root 672 0.0 0.0 0 0 ? - Sep04 0:00 [kpsmoused]
root - 0.0 - - - - S< Sep04 0:00 -
root 685 0.0 0.0 0 0 ? - Sep04 0:00 [kjournald]
root - 0.0 - - - - S< Sep04 0:00 -
root 1666 0.0 0.0 1588 464 ? - Sep04 0:00 udevd
root - 0.0 - - - - S<s Sep04 0:00 -
root 2065 0.0 0.0 0 0 ? - Sep04 0:00 [kjournald]
root - 0.0 - - - - S< Sep04 0:00 -
root 2073 0.0 0.0 0 0 ? - Sep04 0:00 [kjournald]
root - 0.0 - - - - S< Sep04 0:00 -
root 2763 0.0 0.0 2484 928 ? - Sep04 0:00 crond
root - 0.0 - - - - Ss Sep04 0:00 -
root 2779 0.0 0.0 1532 408 tty2 - Sep04 0:00
/sbin/mingetty tty2
root - 0.0 - - - - Ss+ Sep04 0:00 -
root 2780 0.0 0.0 1536 416 tty3 - Sep04 0:00
/sbin/mingetty tty3
root - 0.0 - - - - Ss+ Sep04 0:00 -
root 4495 0.0 0.0 2904 1236 ? - Sep04 0:00 login -- root
root - 0.0 - - - - Ss Sep04 0:00 -
root 6582 0.0 0.0 0 0 ? - Sep04 0:00 [reiserfs/0]
root - 0.0 - - - - S< Sep04 0:00 -
root 7406 0.0 0.0 2344 1380 tty1 - Sep06 0:00 -bash
root - 0.0 - - - - Ss+ Sep06 0:00 -
root 8158 0.0 0.0 4076 1132 ? - Sep06 0:01 /usr/sbin/sshd
root - 0.0 - - - - Ss Sep06 0:01 -
root 10342 0.0 0.0 2244 608 ? - Sep08 0:03 syslogd -m 0
root - 0.0 - - - - Ss Sep08 0:03 -
root 10347 0.0 0.0 1672 384 ? - Sep08 0:00 klogd -x
root - 0.0 - - - - Ss Sep08 0:00 -
root 16078 0.0 0.0 8040 2240 ? - Sep10 0:00 sshd: root@pts/0
root - 0.0 - - - - Ss Sep10 0:00 -
root 16089 0.0 0.0 2516 1256 pts/0 - Sep10 0:00 -bash
root - 0.0 - - - - Ss Sep10 0:00 -
nscd 24962 0.0 0.0 106476 840 ? - 03:43 0:00 /usr/sbin/nscd
nscd - 0.0 - - - - Ssl 03:43 0:00 -
nscd - 0.0 - - - - Ssl 03:43 0:00 -
nscd - 0.0 - - - - Ssl 03:43 0:00 -
nscd - 0.0 - - - - Ssl 03:43 0:00 -
nscd - 0.0 - - - - Ssl 03:43 0:00 -
nscd - 0.0 - - - - Ssl 03:43 0:00 -
nscd - 0.0 - - - - Ssl 03:43 0:00 -
root 24972 0.0 0.0 2548 724 pts/0 - 03:44 0:00 ps auxwm
root - 0.0 - - - - R+ 03:44 0:00 -
> echo t > /proc/sysrq-trigger
> dmesg -s 1000000 > foo
>
> then edit foo, search for the process in D state (look for " D ") and send
> that process's backtrace record.
This does only make sense if I have found the offender? I tried it,
nothing unusual and almost no output.
Is there any way to track it down? Just hints in the right direction
are highly appreciated!
Thanks,
M.
next prev parent reply other threads:[~2006-09-08 21:03 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-09-07 18:08 Metathronius Galabant
2006-09-08 0:53 ` Andrew Morton
2006-09-08 21:03 ` Metathronius Galabant [this message]
2006-09-08 22:47 ` Pavel Machek
2006-09-11 11:05 ` Metathronius Galabant
2006-09-11 11:49 ` Metathronius Galabant
2006-09-11 12:32 ` Vitezslav Samel
-- strict thread matches above, loose matches on Subject: below --
2006-09-07 18:04 Metathronius Galabant
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1b270aae0609081403u11b76ae9v72ad933475a2319f@mail.gmail.com \
--to=m.galabant@googlemail.com \
--cc=akpm@osdl.org \
--cc=linux-kernel@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®