Date: Mon, 05 May 2008 08:41:26 +0300 From: Oleksandr Samoylyk <oleksandr@samoylyk.sumy.ua> To: freebsd-net@freebsd.org Cc: Jack Vogel <jfvogel@gmail.com>, mike@sentex.net Subject: Re: Troubles with em on FreeBSD 7 Message-ID: <481E9E06.3040603@samoylyk.sumy.ua> In-Reply-To: <2a41acea0805041532le60cc9cybf22887e9fdcb3f9@mail.gmail.com> References: <481C84B7.6020205@samoylyk.sumy.ua> <bibs14perk47snkpg8fu691kjkc42gh6m2@4ax.com> <481E338D.6040706@samoylyk.sumy.ua> <2a41acea0805041529j5d4dd2f7x5e07a8d6d2eb89b6@mail.gmail.com> <2a41acea0805041532le60cc9cybf22887e9fdcb3f9@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Jack Vogel wrote:
> Oh, I just had a thought, increase the RX processing limit,
> that only allows you to process 100 packets in one pass.
>
> First change it to 250 and see what it does, you might
> also set it to -1 which will allow you to process til you
> drain the ring, the risk is that you cause other problems
> by doing that, but heck at this point anything is worth
> trying, right?
>
last pid: 55326; load averages: 1.40, 1.24, 1.05
up 0+06:56:52 08:32:41
83 processes: 10 running, 60 sleeping, 13 waiting
CPU states: 0.5% user, 0.0% nice, 10.5% system, 1.1% interrupt, 87.9%
idle
Mem: 42M Active, 265M Inact, 272M Wired, 244K Cache, 214M Buf, 7338M Free
Swap: 4096M Total, 4096M Free
PID USERNAME THR PRI NICE SIZE RES STATE C TIME WCPU COMMAND
12 root 1 171 ki31 0K 16K CPU6 6 413:05 100.00%
idle: cpu6
16 root 1 171 ki31 0K 16K RUN 2 401:28 100.00%
idle: cpu2
15 root 1 171 ki31 0K 16K CPU3 3 402:45 96.09% idle:
cpu3
17 root 1 171 ki31 0K 16K CPU1 1 398:26 92.68% idle:
cpu1
14 root 1 171 ki31 0K 16K CPU4 4 401:24 90.28% idle:
cpu4
18 root 1 171 ki31 0K 16K CPU0 0 382:00 89.26% idle:
cpu0
13 root 1 171 ki31 0K 16K CPU5 5 392:57 81.40% idle:
cpu5
11 root 1 171 ki31 0K 16K CPU7 7 378:44 59.47% idle:
cpu7
30 root 1 -68 - 0K 16K CPU7 7 37:24 44.78% em1 taskq
31 root 1 -68 - 0K 16K - 5 23:00 18.26% em2 taskq
19 root 1 -44 - 0K 16K WAIT 3 35:45 14.60% swi1: net
32 root 1 -68 - 0K 16K - 0 19:51 7.08% em3 taskq
722 root 2 5 0 68280K 21652K select 2 0:00 1.46% mpd5
20 root 1 -32 - 0K 16K WAIT 2 3:56 0.00% swi4:
clock
29 root 1 -68 - 0K 16K - 6 2:53 0.00% em0 taskq
22 root 1 44 - 0K 16K - 1 0:49 0.00% yarrow
I've added more NICs.
The thing is that the most busiest interface is em3 (uplink) and it's
just 7%.
# netstat -I em0 -w 1 (just for the way back traffic)
input (em0) output
packets errs bytes packets errs bytes colls
13 0 2344 8408 0 4290811 0
20 0 2232 8147 0 4179926 0
12 0 1358 8204 0 3984088 0
33 0 5660 7982 0 3944704 0
25 0 3165 8183 0 4226968 0
10 0 1012 8620 0 4288722 0
25 0 2822 8283 0 4220019 0
^C
# netstat -I em1 -w 1
input (em1) output
packets errs bytes packets errs bytes colls
7137 0 2618138 0 0 0 0
6772 0 2565885 0 0 0 0
6982 0 2430458 0 0 0 0
6758 0 2459663 0 0 0 0
7036 0 2665113 0 0 0 0
^C
# netstat -I em2 -w 1
input (em2) output
packets errs bytes packets errs bytes colls
1237 0 327276 0 0 0 0
1158 0 328459 0 0 0 0
1243 0 346992 0 0 0 0
1076 0 323918 0 0 0 0
1242 0 363008 0 0 0 0
^C
# netstat -I em3 -w 1
input (em3) output
packets errs bytes packets errs bytes colls
5681 0 3822499 6233 0 2443102 0
5958 0 4233009 6692 0 2354643 0
6202 0 4296877 6432 0 2455513 0
6117 0 4146643 6419 0 2327346 0
6324 0 4128040 6695 0 2468243 0
5856 0 4052255 6252 0 2445518 0
^C
Why so?
--
Oleksandr Samoylyk
OVS-RIPE
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?481E9E06.3040603>