Re: CBWFQ in the real life

From: Joe Chang (changjoe@earthlink.net)
Date: Thu Jun 10 2004 - 11:43:22 GMT-3


In a sense the interface does want you want already, without modification.
If the only active egress flow is C then there will be no contention for the
interface, and that flow will receive up to the full bandwidth
(police-limited to 2 mbps, in this case.)

My rough guess is that a straightforward config of CBWFQ for three classes
will satisfy your requirements. However, LLQ is probably your best option
because you need the priority queue for voice traffic.

----- Original Message -----
From: <roih@012.net.il>
To: <ccielab@groupstudy.com>
Sent: Thursday, June 10, 2004 6:27 AM
Subject: CBWFQ in the real life

> hi group.
>
> I need help in CBWFQ case in the real world.
> here it goes:
>
>
> 3 classes A,B,C the first is voice.
>
> I have fast Ethernet interface and I need the limit it to 2M (that&apos;s
what
> the
> customer wants to pay).
> in 2M I have the police the traffic the A and B will get
> 0.5M each and C the other 1M.
> Now for the problem: I want to be able the allocate the BW of class A if
no
> packets of voice are arriving and use it
> in class B, the same case from B to C.
> So, my config file have to give me the option to use the entire BW even if
A
> and B are empty. I&apos;ve tried it in few ways (also with the new feature
"two-
> rate
> policer" but no success.
> The BW remaining commands cant help here because it define the way the
> Allocate the left BW after the classes used the BW. My challenge is to
"leak"
> BW from class to class and not from the remaining "bank" to the classes
>
>
> Experts, Any idea??
>
> p.s the version can be 12.3...
>
> Roi
>
> _______________________________________________________________________
> Please help support GroupStudy by purchasing your study materials from:
> http://shop.groupstudy.com
>
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html



This archive was generated by hypermail 2.1.4 : Sat Jul 03 2004 - 19:40:37 GMT-3