Re: CAR configuration parameters - IE-LAB17

From: Alan Ewer (acewer64@gmail.com)
Date: Sun Jul 29 2007 - 07:55:40 ART


Hi Ryan.
nope doesnt expain what i'm after...theres a lot of particulary vague doco
on CAR, what i am looking for is how to answer the question

Do not allow allow traffic to exceed <somevalue> (ie 500000 (bps)

This implies policing and becasue of the CAR/Police BC value, the value you
put in could be exceeded because CAR is a limiter over an extended period of
time. ?? not a strict limiter on bandwidth ...

So I am now in the dark, not only as to how IE come up with their answer but
the whole world of CAR/policing (CBWFQ) is a mystery.

Regds
Alan E

On 7/27/07, Ryan <ryan95842@gmail.com> wrote:
>
> I know this doesn't answer your question, but from command reference,
> there is a formula for configuring the normal and extend burst. I've found
> in "real life" it to be true. Using the 'default' bytes for normal and
> extended burst like the solution shows will seriously degrade the
> performance and it will police at a much lower rate. Using the formula
> though works well. Since this is the lab, I guess it doesn't matter.
>
> I have no idea how they got 496000 either...based on the question and the
> formula, I would have put in:
>
> !
> int e 1/0
> rate-limit output access-group 191 500000 93750 187500 conform-action
> transmit exceed-action drop
> !
>
>
> -Ryan
>
>
> Usage Guidelines
>
> Use this command to configure your CAR policy on an interface. To specify
> multiple policies, enter this command once for each policy.
>
> CAR and DCAR can be configured on an interface or subinterface.
>
> Policing Traffic with CAR
>
> CAR embodies a rate-limiting feature for policing traffic. When policing
> traffic with CAR, Cisco recommends the following values for the normal and
> extended burst parameters:
>
> normal burst = configured rate * (1 byte)/(8 bits) * 1.5 seconds
>
> extended burst = 2 * normal burst
>
> With the listed choices for parameters, extensive test results have shown
> CAR to achieve the configured rate. If the burst values are too low, then
> the achieved rate is often much lower than the configured rate.
>
>
>
>
>
>
>
>
>
> On 7/26/07, Alan Ewer <acewer64@gmail.com> wrote:
>
> > Hi All
> > apologies for asking this here if it has already been answered elsewhere
> > ( i
> > didnt find it if it has .)
> >
> > I am struggling with CAR. Question reads like this... Configure CAR on
> > an
> > interface to not exceed 500kbps outbound..
> >
> > cool i think to myself !!
> >
> > The command syntax is :
> > rate-limit input|output access-group <name> <sustained-rate in bps>
> > <normal
> > burst BYTES> <excess burst BYTES> conform-action transmit exceed-action
> >
> > drop
> >
> > OK this is applied to an ethernet interface on a 3640 and has nothing to
> > do
> > with time-intervals <true/false ???>
> >
> > the answer in the workbook is
> >
> > int e 1/0
> > rate-limit out put access-group 191 496000 4000 4000 conform-action
> > transmit exceed-action drop
> >
> >
> >
> > Question 1 How does the 496000 get derived
> > Question 2 what and where does 4000 bytes normal burst get derived..
> > what
> > are the considering factors here ??
> > Question 3 what difference does the excess burst have if the
> > exceed-action
> > is "drop" ..... i
> >
> > If anyone could point me in the right direction.. i would me grateful...
> > I have already read the QOS guide and the configuration guide and they
> > dont
> > seem to be helping me with this concept !
> >
> > I dont get how 496000 bits per second + 4000 bytes per second = 500000
> > bits
> > per second
> >
> > Regds
> > AlanE
> > Brisbane Oztralia
> >
> > _______________________________________________________________________
> > Subscription information may be found at:
> > http://www.groupstudy.com/list/CCIELab.html



This archive was generated by hypermail 2.1.4 : Sat Aug 18 2007 - 08:17:42 ART