Re: Legacy adaptive FRTS question

Hi,

Here is the problem that many people face understanding the shaping
algorithm. The fact is that "sending" rate is an averaged "fictious"
value. The actual "real" variables are Bc and Tc - shaper sends at
most Bc bits every Tc milliseconds. Every burst is being sent at the
interface access rate (physical rate). Effectively, CIR=Bc/Tc is the
value that appears as a result of averaging across multiple Tc, not
the "real" sending rate.

Now for adaptive shaping. When Frame-Relay layer detects congestion
via some means (signaling or interface congestion) it needs to bring
the "average" CIR to MinCIR. The are two ways to do that per the
formula CIR=Bc/Tc

1) Shrink Bc in size
2) Increase Tc duration

Most of the times it makes sense to decrease Bc (not messing with
latency), which FRTS does using a back-off algorithm. Upon every
congestion notification message received, Bc is decremented until it
is small enough to safisfy MinCIR=Bc/Tc. By default, MinCIR=CIR/2 and
thus FRTS will shrink the burst size to half of the configured value
while responding to congestion.

Once again, remember, even if you set frame-relay traffic rate with
FRTS commands, you effectively encode Bc and Tc which are used by the
shaper's scheduler!

HTH,

-- 
Petr Lapukhov, petr_at_INE.com
CCIE #16379 (R&S/Security/SP/Voice)
Internetwork Expert, Inc.
http://www.INE.com
Toll Free: 877-224-8987
Outside US: 775-826-4344
2009/8/9 Victor <cisco.all.stuff_at_gmail.com>:
> Hello People,
>
> Many thanks for your answer but maybe I'm still confused.
>
> My question is:
>
> In the case of FRTS adaptive, how I should calculate the Bc, using the CIR
> configured or the minCIR??
>
> Bc=CIR*Tc ====== OR ====== Bc=minCIR*Tc ???
>
> I've seen both ways in different documents.
>
> Cheers..... Johan Victor.
>
>
>
> On Sun, Aug 9, 2009 at 4:24 PM, Bryan Bartik <bbartik_at_ipexpert.com> wrote:
>
>> Victor,
>>
>> Sending rate is reduced to the mincir. Check this link out:
>>
>>
>> http://www.cisco.com/en/US/docs/ios/12_2t/12_2t4/feature/guide/ft_afrts.html
>>
>>   On Sun, Aug 9, 2009 at 11:42 AM, Victor <cisco.all.stuff_at_gmail.com>wrote:
>>
>>>  Hello Guys,
>>>
>>> Today my head is a mess, I've been re-reading some documentation about the
>>> subject topic and I realized I've a big confusion, since different author
>>> deal with this in a little different way.  My question is:
>>>
>>> Bc calculation for adaptive shaping is based on CIR or MINCIR??
>>>
>>> Thanks in advance for you help.
>>>
>>> Cheers.... Johan Victor
>>>
>>>
>>> Blogs and organic groups at http://www.ccie.net
>>>
>>> _______________________________________________________________________
>>> Subscription information may be found at:
>>> http://www.groupstudy.com/list/CCIELab.html
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>
>>
>> --
>> Bryan Bartik
>> CCIE #23707 (R&S), CCNP
>> Sr. Support Engineer - IPexpert, Inc.
>> URL: http://www.IPexpert.com <http://www.ipexpert.com/>
>
>
> 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