Re: Frame Relay Traffic Shaping .. Karl Soli Example 5-17

From: Shanky (shankyz@gmail.com)
Date: Sat Jul 16 2005 - 09:47:38 GMT-3


Hi Chris,
 I think there is no point in configuring the CIR of the central side =
Access rate of the link to the FR cloud and thats what Mr Solie seems to
suggest. CIR and AR are 2 seperate things and I could negotiate and agree to
a CIR for a PVC depending upon the remote & Central side links AR. Now , may
be I could have lets say 20 remote locations with 64 k each and at central
location 1 T1, so in this case also I'd be negotiating for a CIR of 64k max
for each of the PVC and not T1 rate for the central side.
 so , in my opinion AR is not equal to CIR.
 Regards
 Shanky

 On 7/16/05, Chris Lewis (chrlewis) <chrlewis@cisco.com> wrote:
>
> He is setting the CIR to port speed. That is the easy part :)
>
> Is this correct? Well, I guess it depends.
>
> He has a central site connected via T1 to a frame cloud. By setting CIR
> to T1 speed and mincir to 32k (and enabling adaptive shaping), the
> central site port will transmit at a speed between 32k and 1.5M
> depending on the presence of BECNs from the network. The remote ends
> only have 64K lines, so sustained transfers from the central site to a
> remote site will have to throttle down to this eventually. From that
> perspective it is a bit pointless setting the CIR value so high on the
> central site. Setting it to the remote port speed would mean fewer
> dropped packets and no BECNs, a better choice if you have layer 3 QoS
> mechanisms and want layer 3 to make drop decisions rather than the less
> application aware layer 2.
>
> However if at some point in the future there is a larger site added to
> the network with a T1 access, you might want to have the CIR set to T1
> on the central site so it can take advantage of the one site that can
> accept T1 traffic rate, and accept packets to the slower sites
> potentially being dropped and BECNs slowing down the transfer to those
> sites.
>
> Chris
>
> -----Original Message-----
> From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
> Shanky
> Sent: Saturday, July 16, 2005 1:53 AM
> To: ccielab@groupstudy.com
> Subject: Frame Relay Traffic Shaping .. Karl Soli Example 5-17
>
> Hi,
> In the example, he is setting up the CIR to be the Phy Port speed ? Is
> it correct ?
> Regards
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html



This archive was generated by hypermail 2.1.4 : Sun Sep 04 2005 - 17:00:30 GMT-3