From: Alex Tomas <bzzz@tmi.comex.ru>
To: linux-kernel@vger.kernel.org
Subject: [ANNOUNCE] fast access to process list
Date: 26 Dec 2002 00:15:56 +0300 [thread overview]
Message-ID: <m3of796bqr.fsf@lexa.home.net> (raw)
Hello!
Look how long is following command on system with 3000 processes.
root@zefir:~# time bash -c 'let i=0; while let "i<10"; do ps axl; let "i=i+1"; done>/dev/null'
real 0m10.380s
user 0m5.511s
sys 0m4.870s
Ie, ten ps'es eat 10 secs and system had no active processes at all.
I really don't like it. Therefore, I started to implement own chardev driver
with own tool (I call it 'fast ps'). Look how long it needs:
root@zefir:~/fastps# time bash -c 'let i=0; while let "i<10"; do ./fps l; let "i=i+1"; done>/dev/null'
real 0m0.992s
user 0m0.620s
sys 0m0.372s
Of course, the tool isn't ready yet, but things change ... Here is output of fast ps:
[root@lexx fastps]# ./fps -l
F UID PID PPID PRI NI VSIZE RSS WCHAN STAT TTY TIME COMMAND
100 0 1 0 29 0 1372 424 do_sel S ? 0:08 init [5]
040 0 2 1 30 0 0 0 contex S ? 0:07 keventd
040 0 3 1 30 0 0 0 apm_ma S ? 0:00 kapmd
040 0 4 1 40 19 0 0 ksofti SN ? 0:03 ksoftirqd
840 0 5 1 30 0 0 0 kswapd S ? 0:05 kswapd
040 0 6 1 30 0 0 0 bdflus S ? 0:00 bdflush
040 0 7 1 30 0 0 0 kupdat S ? 0:00 kupdated
040 0 11 1 30 0 0 0 end S ? 0:01 kjournald
040 0 90 1 30 0 0 0 end S ? 0:00 khubd
040 0 191 1 30 0 0 0 end S ? 0:03 kjournald
040 0 462 1 30 0 1428 492 do_sel S ? 0:00 syslogd -
140 0 467 1 30 0 1364 428 do_sys S ? 0:00 klogd -x
140 32 487 1 30 0 1512 452 do_pol S ? 0:00 portmap
140 0 517 1 30 0 1696 524 do_pol S ? 0:00 rpc.mount
140 0 522 1 30 0 0 0 end S ? 0:00 nfsd
140 0 523 1 30 0 0 0 end S ? 0:01 nfsd
040 0 524 1 30 0 0 0 end S ? 0:00 lockd
040 0 525 524 30 0 0 0 end S ? 0:00 rpciod
140 0 636 1 30 0 1488 440 do_sel S ? 0:00 /sbin/car
140 0 654 1 29 0 1360 412 do_sel S ? 0:00 /usr/sbin
140 0 723 1 29 0 2620 652 do_sel S ? 0:00 /usr/sbin
140 0 743 1 30 0 1536 536 nanosl S ? 0:00 crond
040 0 761 1 30 0 1404 464 nanosl S ? 0:00 /usr/sbin
100 0 778 1 30 0 1344 336 read_c S tty1 0:00 /sbin/min
100 0 779 1 30 0 3212 644 do_sel S ? 0:00 /usr/X11R
040 0 5499 1 30 0 0 0 end S ? 0:00 kjournald
100 0 20899 779 38 0 25260 10052 - R ? 8:57 /usr/X11R
At this moment, I have patches against 2.4.20 and 2.5.53. You may find tarball
at http://tmi.comex.ru/fastps.tgz.
I would love to hear any comments/suggestions!
with best regards, Alex
next reply other threads:[~2002-12-25 21:15 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-12-25 21:15 Alex Tomas [this message]
2002-12-31 8:06 Alex Tomas
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=m3of796bqr.fsf@lexa.home.net \
--to=bzzz@tmi.comex.ru \
--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®