Re: QOS Bc Be and Tc

Karim,

These are all correct, nice and perfectly explained but I usually don't read
through tens of lines :) instead I'd rather take a look at a simple diagram
to see "WHERE" each one of these things are.

If you're still unsure, drop me a line I can show you where they are...

Kambiz Agahian

CCIE Instructor/Consultant
M.Eng Telecom, CCIE# 25341, CCSI# 33326, MCSE, MCSA

On Wed, Aug 18, 2010 at 1:45 PM, karim jamali <karim.jamali_at_gmail.com>wrote:

> Hi Experts,
>
> I believe I have something to correct regarding the Be concept I spoke of.
> Well let me put it this way..Let us for say assume that I have a Bc
> (CIR/Tc)
> 8000 bits (these are the bits I can send every interval let us suppose Tc
> =125 ms. Thus In one sec I can send 64000 bits and in 30 seconds for
> example
> I can send 64000*30=1920000 bits. Let us suppose for instance, that I
> didn't
> send for a 30 second interval, thus I have wasted this amount of bits
> (1,920,000) bits, and now a new interval has come Tc. How much I can send
> in
> this interval? Well just by looking @ it I would say that I can send my Bc(
> dedicated for this Tc) as well as my account (1,920,000) which I have
> wasted
> over my last 30 second interval. However let us suppose that the AR of the
> interface is 1024 kbps. Thus the maximum number of bits I am allowed to
> transfer in terms of the interface capability is 1024kbps/8=128000 bit/Tc
> (64000 already the Tc for the interval) and a maximum I can add of 64000
> bits.(Compare this to my account 1,920,000 bits). Thus my account will be
> re-deposited within 30 seconds where every Tc I can add 64000 bits to my Bc
> as a Be value. Note I think if the AR limitation didn't exist for instance
> in a scenario, the account will be re-deposited fully in a single Tc.
>
> I would like to thank Carlos G Mendioroz for shedding the light on this,
> and
> Kambiz Aghian who explained it in the link sent by Carlos.
>
> Thank You:)
>
> Best Regards,
>
> KJ
>
>
> On Wed, Aug 18, 2010 at 11:11 PM, Carlos G Mendioroz <tron_at_huapi.ba.ar
> >wrote:
>
> > Take a look at
> >
> > http://news.groupstudy.com/archives/ccielab/200207/msg01417.html
> >
> > or
> >
> >
> >
>
> http://www.2bccie.com/cisco-certification/ccie-shaping-and-calculating-be-val
> ue-for-a-30-second-burst.html<http://www.2bccie.com/cisco-certification/ccie-shaping-and-calculating-be-val%0Aue-for-a-30-second-burst.html>
> >
> > Have your own conclusions...
> > -Carlos
> >
> > karim jamali @ 18/8/2010 17:02 -0300 dixit:
> >
> >> Hi Carlos,
> >>
> >> You may be right..i will check the archives. However I also checked
> Petr's
> >> document and found the following formula:
> >> PIR = CIR*(1+Be/Bc). By looking @ this formula one can also see the
> >> relationship between PIR/Be.
> >>
> >>
> >> http://blog.ine.com/2008/08/26/understanding-the-shape-peak-command/
> >>
> >> I will be checking, please if you have a good reference will appreciate
> >> it!
> >>
> >> Thanks man:)
> >>
> >>
> >>
> >> On Wed, Aug 18, 2010 at 10:54 PM, Carlos G Mendioroz
> <tron_at_huapi.ba.ar<mailto:
> >> tron_at_huapi.ba.ar>> wrote:
> >>
> >> Please check your knowledge with a simple test.
> >> Search GS archives for discussions on this. Be can actually be
> >> transmitted over many Tcs.
> >>
> >>
> >> You can not enforce a PIR with Tc,Bc and Be. You need PIR to be
> >> explicit.
> >>
> >> -Carlos
> >>
> >> karim jamali @ 18/8/2010 16:50 -0300 dixit:
> >>
> >> Hi Carlos,
> >>
> >> I know Bc/Be are both amounts of data..But as per my knowledge
> >> Bc/Be does need to happen in a single Tc and they relate
> >> directly to CIR/PIR; they are in fact the amount of data that
> >> can be sent during one Tc interval.
> >>
> >>
> >>
> http://qospf.wordpress.com/2008/06/01/ccie-cert-guide-notes-qos-part-iv/
> >> *Be = Tc * (PIR CIR);
> >>
> >> *The maximum Be is in fact found by replacing the PIR by the AR
> >> (Access Rate of the interface).
> >>
> >> Best Regards,
> >>
> >> On Wed, Aug 18, 2010 at 10:41 PM, Carlos G Mendioroz
> >> <tron_at_huapi.ba.ar <mailto:tron_at_huapi.ba.ar>
> >> <mailto:tron_at_huapi.ba.ar <mailto:tron_at_huapi.ba.ar>>> wrote:
> >>
> >> Karim,
> >> Be does not need to happen in one Tc, as your message implies.
> >> Also, Be is not a rate, is just an ammount of burst.
> >> That's the difference between one rate and two rate policers,
> >> and so PIR is not directly related to Be.
> >>
> >> -Carlos
> >>
> >> karim jamali @ 18/8/2010 16:20 -0300 dixit:
> >>
> >> Hi Masroor,
> >>
> >> Well the simplest and easiest way to think about it
> >> (whether with
> >> policing/shaping) is that you are connected via a
> >> Fastethernet
> >> interface to
> >> a provider (however most probably you won't have 100 Mbps
> >> bought
> >> from the
> >> provider), and one thing about interfaces they are either
> >> ON or
> >> OFF. So you
> >> can tell it please only send 2 Mbps or 3Mbps..Thus the
> >> only way
> >> to control
> >> is to turn it ON/OFF in such a manner that you will be
> >> sending
> >> the rate you
> >> already bought.
> >>
> >> Thus the CIR is what you already agreed upon with the
> >> Provider.
> >> Bc is the
> >> same as the CIR however within a smaller time frame Tc.
> >> Suppose
> >> for instance
> >> you are given 2Mbps (CIR) from a provider, and your Tc is
> >> 125
> >> ms..Your
> >> router will divide this one sec into 8 intervals (each a
> Tc
> >> long) and in
> >> that Tc you will be only sending Bc (what is allowed to
> >> send in
> >> one Tc).
> >> However note that to send the Bc you will require only a
> >> part of
> >> the Tc for
> >> instance half. Thus what is actually happening is you are
> >> sending for half a
> >> Tc, and putting your interface off for another half of the
> >> Tc
> >> with the aim
> >> of getting an average of 2Mbps.
> >>
> >> Calculation Wise: 2Mbps/1 sec = Bc/Tc=CIR/1sec thus you
> can
> >> calculate the
> >> Bc.
> >> Then check how much time it takes you to send the Bc
> >> (based ont
> >> he interface
> >> speed).
> >> time required= Bc/(Access Rate of interface) = (this time
> >> will
> >> be definitely
> >> less than Tc)
> >>
> >> Thus in every Tc interval you will be ON for this time
> >> required/off for the
> >> remaining time. Note that for VOIP applications it is
> >> advised
> >> that the Tc be
> >> small so that your interface doesn't turn off for a long
> >> time
> >> and break the
> >> 150ms rule.
> >>
> >> Regarding Be, is in fact what happens if you exceed the
> rate
> >> that your ISP
> >> has given you..It happens in a way that suppose for one
> >> of the
> >> Tc intervals,
> >> you undersubscribed your utilization, i.e. you didn't
> >> reach your
> >> Bc limit,
> >> the provider might be kind enough to keep you some
> >> balance, and
> >> in another
> >> time you may be allowed to burst above Bc. Thus in one Tc
> it
> >> might be
> >> possible that you send (Bc + Be) the Be being the access
> >> in your
> >> account.
> >> Same formulas/same concept with CIR but this time with
> PIR.
> >> Supposing that
> >> the ISP tells you I am comitted to serve you 2Mbps but at
> >> times
> >> you may be
> >> allowed to get a peak PIR 3Mbps.
> >>
> >>
> >> Hope this helps..I guess for more information you can
> >> refer to
> >> INE's blog as
> >> Petr's posts are great.
> >> http://blog.ine.com/page/3/?s=QOS+Shaping
> >>
> >> Best Regards,
> >>
> >> On Wed, Aug 18, 2010 at 9:52 PM, masroor ali
> >> <masror.ali_at_gmail.com <mailto:masror.ali_at_gmail.com>
> >> <mailto:masror.ali_at_gmail.com <mailto:masror.ali_at_gmail.com>>>
> >> wrote:
> >>
> >> hi,
> >>
> >> Can anyone explain in easy wordings that what is Bc
> >> Be and Tc
> >>
> >> --
> >> Regards,
> >> Masroor Ali
> >>
> >>
> >> Blogs and organic groups at http://www.ccie.net
> >>
> >>
> >> _______________________________________________________________________
> >> Subscription information may be found at:
> >> http://www.groupstudy.com/list/CCIELab.html
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> -- Carlos G Mendioroz <tron_at_huapi.ba.ar
> >> <mailto:tron_at_huapi.ba.ar> <mailto:tron_at_huapi.ba.ar
> >>
> >> <mailto:tron_at_huapi.ba.ar>>>
> >> LW7 EQI Argentina
> >>
> >>
> >>
> >>
> >> -- KJ
> >>
> >>
> >> -- Carlos G Mendioroz <tron_at_huapi.ba.ar <mailto:
> tron_at_huapi.ba.ar
> >> >>
> >> LW7 EQI Argentina
> >>
> >>
> >>
> >>
> >> --
> >> KJ
> >>
> >
> > --
> > Carlos G Mendioroz <tron_at_huapi.ba.ar> LW7 EQI Argentina
> >
>
>
>
> --
> KJ
>
>
> Blogs and organic groups at http://www.ccie.net
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html

Blogs and organic groups at http://www.ccie.net
Received on Wed Aug 18 2010 - 14:11:29 ART