From: eicc tester (reto_ccie@yahoo.com)
Date: Thu Apr 26 2007 - 09:05:02 ART
In your example, that meam that every dynamic queue of WFQ are subjet to WRED, remember that WRED is the process of delete packets of those queues when is near to be full (the ranges are configurable), ir order to make TCP windows shorter.
"Filyurin, Yan" <yan.filyurin@eds.com> wrote: So does it mean that each of the 64 queues will be subject to WRED, or that WRED will take place before packets are placed into dynamic queues. This is what it sounds according to CISCO documentation.
FOr example what would happen if I were to do something like that:
policy-map mypolicy
class class-default
fair-queue
random-detect
queue-limit 100
?
Thank you
---------------------------------
From: eicc tester [mailto:reto_ccie@yahoo.com]
Sent: Tuesday, April 24, 2007 6:28 PM
To: Filyurin, Yan; ccielab@groupstudy.com
Subject: Re: WFQ and WRED
Remember that queues in WFQ are automatically create and the maximun default is 64, then for the life of the queue (for temporary flow), all WRED setting are valid.
"Filyurin, Yan" <yan.filyurin@eds.com> wrote: Hello esteemed members group study. I have a rather stupid QOS question
that has to do with using WFQ and WRED simultaneously. If were to take
an interface or a class in a policy map to that I would apply fair-queue
configuration with default settings and then also random-detect with
default or non-default setting. Assuming that default WFQ settings are
64 queues, would it mean that by applying WRED, I would effective give
random detect capability to each one of those queues? Does it sound
right and would it ever make sense to do it?
Thank you!
This archive was generated by hypermail 2.1.4 : Tue May 01 2007 - 08:28:38 ART