From: "Radosław Łoboda" <p0wer@tl.krakow.pl>
To: lartc@vger.kernel.org
Subject: Re: [LARTC] HTB equivalent for 'bounded' and 'isolated' in CBQ
Date: Thu, 29 Aug 2002 06:26:28 +0000 [thread overview]
Message-ID: <marc-lartc-103060211111337@msgid-missing> (raw)
In-Reply-To: <marc-lartc-102953224700762@msgid-missing>
On Tue, 20 Aug 2002, Amit Kucheria wrote:
> On Fri, 16 Aug 2002, Stef Coene wrote:
>
> >> The 'ceil' parameter allows us to implement a part of the 'isolated'
> >> definition above. By setting it equal to 'rate', the class does not
> >> borrow bandwidth, but what will stop other classes from borrowing its
> >> unused bandwidth?
> >
> > Providing no ceil parameter means rate = ceil. But this can not be
> > used to implement isolated like in cbq. Rate = ceil means the class
> > can not use more bandwidth then it's rate. Isolated in cbq means other
> > classes can not borrow bandwidth from the class and that's not the
> > same. There is no way you can implement isolated with htb.
> >
> > But with htb you can do something like this :
> >
> > Total : 100
> > class1 rate 20 ceil 20
> > class2 rate 40 ceil 80
> > class3 rate 40 ceil 80
> >
> > class 1 is isolated like in cbq. It can not use more then it's
> > rate/ceil and class2 and class3 will never use bandwidth from class1,
> > only from each other. Just like the definition of isolated :)
>
> What happens when there is no traffic in class 1? Will class 2 and 3
> share class 1's bandwidth then?
>
> The reason I am asking these questions is because I am trying to see if
> HTB can be used in place of CBQ in Diffserv. Has anybody tried something
> like this?
It works fine for me, although I haven't used CBQ previously so I can't
comapre them. It just works for me, except the delays in interactive class
are too long, even when the prio is 0 (highest) and the rate and ceil are
maximum. Maybe CBQ is better in reducing the delays... Anybody solved this
problem?
--
##########################################
# | p0wer | #
# __ | GG#1877248 | #
# (oo) | p0wer@bojko.eu.org | #
# / \/ \ Go away or I will replace you #
# `V__V' with a very small shell script. #
##########################################
_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
next prev parent reply other threads:[~2002-08-29 6:26 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-08-16 21:09 [LARTC] HTB equivalent for 'bounded' and 'isolated' in CBQ Amit Kucheria
2002-08-18 20:01 ` Stef Coene
2002-08-21 3:32 ` Amit Kucheria
2002-08-21 17:48 ` Stef Coene
2002-08-29 6:26 ` Radosław Łoboda [this message]
2002-08-29 13:08 ` Stef Coene
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=marc-lartc-103060211111337@msgid-missing \
--to=p0wer@tl.krakow.pl \
--cc=lartc@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 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.