From: Martin Dean (mdean@diatac.co.uk)
Date: Mon Jul 03 2006 - 07:39:47 ART
Great info.
Do you happen to know how the AQD is calculated?
There is a constant n (default=9) when used to calculate the AQD, which is
based on the old AQD and the real FIFO Q size.
But how often is this calculated?
All the formula I have seen just shows Qavg(t+1) without declaring the
interval period.
Any ideas of the interval period?
Martin
----- Original Message -----
From: "Petr Lapukhov" <petr@internetworkexpert.com>
To: "Victor Cappuccio" <cvictor@protokolgroup.com>
Cc: "Cisco certification" <ccielab@groupstudy.com>
Sent: Monday, July 03, 2006 7:45 AM
Subject: Re: MDP .- WRED
> Oh, I see your point now :)
>
> Now, let's take a look how it works:
>
> MDP is just a *probablity* of drop, when you reach the upper limit.
>
> RED tracks "smoothed" average queue depth (AQD). Queue is
> FIFO, so scheduler basically just takes the first packet and sends
> it, _or_ drops it, depending on AQD and thresholds.
>
> When AQD reaches low-threshold, you have drop probability of zero.
> That is, no packets are dropped here, first packet in FIFO queue
> is always sent.
>
> When AQD grows from low to high threshold, drop probability is
> linearly interpolated:
>
> drop_prob = (AQD-min_thresh)/(max_thresh-min_thresh)*MDP
>
> Finally, when you reach high threshold, approx every 1 of MDP
> packets in queue head will be dropped, and every other is sent.
> Above that threshold, every packet is dropped by scheduler
>
> You see, MDP is a value to calculate drop_prob, and make a
> "decision" for scheduler - send upper packet in FIFO queue, or
> drop it.
>
> HTH
> --
> Petr Lapukhov, CCIE #16379
> petr@internetworkexpert.com
>
> Internetwork Expert, Inc.
> http://www.InternetworkExpert.com
> Toll Free: 877-224-8987
> Outside US: 775-826-4344
>
> 2006/7/3, Victor Cappuccio <cvictor@protokolgroup.com>:
> >
> > Petr, that was very informative.
> >
> >
> >
> > But I'm looking for the formula here.
> >
> >
> >
> > So if the Theory Says that the MPD is the fraction of packets drops, how
> > many packets are dropped using that Config that you posed ? 512 of 20?
> >
> > 0,0390625 when the Average Q Depth is 20?
> >
> > Thanks
> >
> >
> >
> >
> >
> > _____
> >
> > De: petrsoft@gmail.com [mailto:petrsoft@gmail.com] En nombre de Petr
> > Lapukhov
> > Enviado el: Lunes, 03 de Julio de 2006 01:36 a.m.
> > Para: Victor Cappuccio
> > CC: Cisco certification
> > Asunto: Re: MDP .- WRED
> >
> >
> >
> > Victor,
> >
> > It may be set separately for precedence or dscp value.
> > There are some basic default, at the same time.
> >
> > e.g.
> >
> > Rack1R6(config-if)#random-detect
> > Rack1R6(config-if)#random-detect precedence 0 10 20 512
> >
> > 0 - precedence value
> > 10 - min thresh
> > 20 - max thresh
> > 512 - MPD
> >
> > --
> > Petr Lapukhov, CCIE #16379
> > petr@internetworkexpert.com
> >
> > Internetwork Expert, Inc.
> > http://www.InternetworkExpert.com
> > Toll Free: 877-224-8987
> > Outside US: 775-826-4344
> >
> > 2006/7/3, Victor Cappuccio <cvictor@protokolgroup.com>:
> >
> > Hi guys,
> >
> > I just can not get what they are trying to say
> >
> >
http://www.cisco.com/en/US/products/sw/iosswrel/ps1831/products_configuratio
> >
> > n_guide_chapter09186a00800c60d9.html#1000915
> >
> > The mark probability denominator is the fraction of packets dropped when
> > the
> > average queue depth is at the maximum threshold. For example, if the
> > denominator is 512, one out of every 512 packets is dropped when the
> > average
> >
> > queue is at the maximum threshold.
> >
> > But then the MAX-TH of the Q depth is set to what number? 512??
> >
> > I think that they forgot to tell that
> >
> > Opinions?
> >
> > Thanks
> > Victor.-
> >
> > _______________________________________________________________________
> > Subscription information may be found at:
> > http://www.groupstudy.com/list/CCIELab.html
> >
> > _______________________________________________________________________
> > Subscription information may be found at:
> > http://www.groupstudy.com/list/CCIELab.html
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
>
>
> --
> No virus found in this incoming message.
> Checked by AVG Free Edition.
> Version: 7.1.394 / Virus Database: 268.9.8/380 - Release Date: 30/06/2006
This archive was generated by hypermail 2.1.4 : Tue Aug 01 2006 - 07:13:46 ART