On Thursday 12 December 2002 09:36, Abraham van der Merwe wrote:
> Hi Stef!
> I think your explanation of "rate of class" in your FAQ is wrong. it
> caught me as well, but from devik's faq page
> (http://luxik.cdi.cz/~devik/qos/htb/htbfaq.htm):
>
> ------------< snip <------< snip <------< snip <------------
> What if sum of child rates is greater than parent rate ?
>
> Then interesting things can happen. Total rate delivered by children can be
> higher that parent's rate (thus its rate is not respected). However when
> sum of actual child rates are under parent's rate then borrowing will occur
> like in regular case.
>
> ------------< snip <------< snip <------< snip <------------
>
> Thus, if sum (rate of childs) > parent rate, then the parent rate is _not
> respected_.
:)
Devik is right, I'm wrong. Do you know why I was convinced that my rule was
OK? Because I always followed my own rules :)
I did some small tests :
ceil rate
1 200 200
\- 10 200 100
\- 100 200 100
\- 101 200 100
\- 20 200 100
Traffic in 100, 101 and 20 : each gets 67 (33%)
Parent rate (10) is not respected otherwise 10 should share 50% with
20.
Traffic in 100 and 20 : each gets 100 (50%)
If you respect "sum (rate of childs) = parent rate" it works like expected :
ceil rate
1 200 200
\- 10 200 100
\- 100 200 50
\- 101 200 50
\- 20 200 100
Traffic in 100, 101 and 20 : 20 gets 100 (50%), 101 50 (25%) and 101 50 (25%).
Traffic in 100 and 20 : each gets 100 (50%)
Conclusion : if you want to know how traffic will be shaped, you have to
follow some basic rules.
I will update docum.org tonight.
Stef
--
[EMAIL PROTECTED]
"Using Linux as bandwidth manager"
http://www.docum.org/
#lartc @ irc.oftc.net
_______________________________________________
LARTC mailing list / [EMAIL PROTECTED]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
