* [LARTC] tc htb stats
@ 2003-02-04 16:10 David DeLauro
2003-02-04 16:20 ` Martin A. Brown
` (2 more replies)
0 siblings, 3 replies; 4+ messages in thread
From: David DeLauro @ 2003-02-04 16:10 UTC (permalink / raw)
To: lartc
I have an htb qdisc setup and it is shaping my traffic wonderfully but
I've just recently started taking a look at the output from the command
tc -s class ls dev eth?
Looking at the "rate" line I thought to myself..."cool I can use this to
graph throughput avg.es without having to worry about the "Sent bytes"
rolling over or being lost when the htb qdisc is recreated" _BUT_ it
seems the rates are inaccurate. When I look at the rate my cisco router
reports for a 5min avg and I look at a calculated 5min average (30 polls
of the tc once every 10sec) for tc they are very different. The cisco is
reported as ~2090kb/s and tc reports my parent as ~251kb/s. Even if I
take all the classes separately and add the averages together I get closer
but not that much closer. Is my math wrong? How does the tc/htb qdisc
figure the avg. rate?
--
David DeLauro
Talk sense to a fool and he calls you foolish. - Euripides
Those who do not think about their own sins make up for it by thinking incessantly about the sins of others. - C. S. Lewis (God in the Dock, "Miserable Offenders," p. 124)
Never undertake anything for which you wouldn't have the courage to ask the blessing of heaven. - G. C. Lichtenberg (1742-1799)
_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
^ permalink raw reply [flat|nested] 4+ messages in thread* Re: [LARTC] tc htb stats
2003-02-04 16:10 [LARTC] tc htb stats David DeLauro
@ 2003-02-04 16:20 ` Martin A. Brown
2003-02-04 16:30 ` David DeLauro
2003-02-04 16:37 ` Stef Coene
2 siblings, 0 replies; 4+ messages in thread
From: Martin A. Brown @ 2003-02-04 16:20 UTC (permalink / raw)
To: lartc
David,
: When I look at the rate my cisco router
: reports for a 5min avg and I look at a calculated 5min average (30 polls
: of the tc once every 10sec) for tc they are very different. The cisco is
: reported as ~2090kb/s and tc reports my parent as ~251kb/s.
: Even if I take all the classes separately and add the averages together
: I get closer but not that much closer. Is my math wrong? How does the
: tc/htb qdisc figure the avg. rate?
$ expr 251 \* 8
2008
Snipped from http://www.docum.org/stef.coene/qos/docs/ :
"When you use the tc command, you have to remember that mb = 1024 kb 1024 * 1024 b = byte/s and that mbit = 1024 kbit = 1024 * 1024 bit bit/s."
Snipped from http://lartc.org/howto/lartc.qdisc.html :
"Just to prevent confusion, tc uses the following rules for bandwith
specification:
mbps = 1024 kbps = 1024 * 1024 bps => byte/s
mbit = 1024 kbit => kilo bit/s.
mb = 1024 kb = 1024 * 1024 b => byte
mbit = 1024 kbit => kilo bit.!
Are you certain your units are the same? Strangely, the tc command line
tool uses kb for kilobytes/second. I think 251 kilobytes-per-second is
quite close to 2090 kilobits-per-second.
Bye for now,
-Martin
--
Martin A. Brown --- SecurePipe, Inc. --- mabrown@securepipe.com
_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [LARTC] tc htb stats
2003-02-04 16:10 [LARTC] tc htb stats David DeLauro
2003-02-04 16:20 ` Martin A. Brown
@ 2003-02-04 16:30 ` David DeLauro
2003-02-04 16:37 ` Stef Coene
2 siblings, 0 replies; 4+ messages in thread
From: David DeLauro @ 2003-02-04 16:30 UTC (permalink / raw)
To: lartc
On Tue, 4 Feb 2003 At 10:20 -0600, mabrown-lartc@securepipe.com Articulated:
>
> Snipped from http://lartc.org/howto/lartc.qdisc.html :
>
> "Just to prevent confusion, tc uses the following rules for bandwith
> specification:
>
> mbps = 1024 kbps = 1024 * 1024 bps => byte/s
Thanks Martin...
I was running under the assuption that "bps" was bits/s and "Bps" was
bytes/sec...
peace
--
David DeLauro
Hata ukinichukia la kweli nitakwambia - Kanga Proverb
Apart from God every activity is merely a passing whiff of insignificance. - Alfred North Whitehead (1861-1947)
Never undertake anything for which you wouldn't have the courage to ask the blessing of heaven. - G. C. Lichtenberg (1742-1799)
_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [LARTC] tc htb stats
2003-02-04 16:10 [LARTC] tc htb stats David DeLauro
2003-02-04 16:20 ` Martin A. Brown
2003-02-04 16:30 ` David DeLauro
@ 2003-02-04 16:37 ` Stef Coene
2 siblings, 0 replies; 4+ messages in thread
From: Stef Coene @ 2003-02-04 16:37 UTC (permalink / raw)
To: lartc
On Tuesday 04 February 2003 17:30, David DeLauro wrote:
> On Tue, 4 Feb 2003 At 10:20 -0600, mabrown-lartc@securepipe.com Articulated:
> > Snipped from http://lartc.org/howto/lartc.qdisc.html :
> >
> > "Just to prevent confusion, tc uses the following rules for bandwith
> > specification:
> >
> > mbps = 1024 kbps = 1024 * 1024 bps => byte/s
>
> Thanks Martin...
>
> I was running under the assuption that "bps" was bits/s and "Bps" was
> bytes/sec...
It's bit when there is bit, otherwise it's byte :)
Stef
--
stef.coene@docum.org
"Using Linux as bandwidth manager"
http://www.docum.org/
#lartc @ irc.oftc.net
_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2003-02-04 16:37 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-02-04 16:10 [LARTC] tc htb stats David DeLauro
2003-02-04 16:20 ` Martin A. Brown
2003-02-04 16:30 ` David DeLauro
2003-02-04 16:37 ` Stef Coene
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.