Re: MQC - FRTS

From: eicc tester (reto_ccie@yahoo.com)
Date: Mon Aug 13 2007 - 11:30:32 ART


This is a good point to discuss, in order to avoid misinterpretation in the lab requirements.
  

Herbert Maosa <asawilunda@googlemail.com> wrote:
    I believe he IS doing FRTS, only that he is using MQC and not legacy FRTS. Using MQC you do not need to specify your traffic shaping parameters with the frame-relay CIR and frame-relay minCIR commands. The CIR becomes what you specify with the shape average|peak command and the minCIR is what you specify with the shape adapative command.
   
  

 
  On 8/13/07, eicc tester <reto_ccie@yahoo.com> wrote: Hi,

In your config really your are not doing FRTS (frame relay traffic shapping) for that you need to configure the command frame-relay cir and mincir under map-class frame relay and active frame-relay traffic shaping on interface.

you configure fragment of 80 under map-class, this mean, that FRF-12, will be fragmenting every 640 bits. this mean about 10 msg.

But if you see the configuration you will see that your MQC shaping give you a TC of 125msg. please see below:

R1#sh policy-map INterface SErial 1/4.601
Serial1/4.601: DLCI 601 -
   Service-policy output: shape_policy_map
     Class-map: class-default (match-any)
     2 packets, 405 bytes
     5 minute offered rate 0 bps, drop rate 0 bps
     Match: any
     Traffic Shaping
          Target/Average Byte Sustain Excess Interval Increment
            Rate Limit bits/int bits/int (ms) (bytes)
           64000/64000 2000 8000 8000 125 1000
         Adapt Queue Packets Bytes Packets Bytes Shaping
       Active Depth Delayed Delayed Active
       BECN 0 6 433 5 353 no

I suggest that you set the bc on shape average comand in order to ensure that you Tc be 10 msg. and agree with fragment use in FRF-12 fragment, to be consistent.

R1(config-pmap-c)#SHape AVerage 64000 640

sh policy-map INterface SErial 1/4.601
Serial1/4.601: DLCI 601 -
   Service-policy output: shape_policy_map
     Class-map: class-default (match-any)
     2 packets, 405 bytes
     5 minute offered rate 0 bps, drop rate 0 bps
     Match: any
     Traffic Shaping
          Target/Average Byte Sustain Excess Interval Increment
            Rate Limit bits/int bits/int (ms) (bytes)
           64000/64000 160 640 640 10 80
         Adapt Queue Packets Bytes Packets Bytes Shaping
       Active Depth Delayed Delayed Active
       BECN 0 6 433 5 353 no

Voice traffic always have guaranted of 32 Kbps.

Ken Young <CiscoKid@ns.sympatico.ca> wrote:
Can anyone confirm my assumption about the following configuration?

- CIR = 64 kbps, miniCIR= 32 kbps.

- A nested LLQ policy which guarantees 32 kbps of low latency bandwidth for
voice (or 'ef' traffic).

*IF* a BECN is received on the PVC; this router will slow its shaping rate
to miniCIR of 32 kbps, because we have a llq configured for 32 kbps of
voice, during a period of congestion it is 'theatrically' possible that ONLY
voice traffic would be transmitted if the voice Q was pushing the full 32
kbps worth of voice traffic.

Am I interrupting that correctly?

class-map voice

match ip dscp ef

!

policy-map llq

class voice

priority 32

!

policy-map shape_policy_map

class class-default

shape average 64000

shape adaptive 32000

service-policy llq

!

map-class frame-relay shape_map_class

frame-relay fragment 80

service-policy output shape_policy_map

!

interface serial0/0

encapsulation frame-relay

!

interface serial0/0.601 point-to-point

ip address 192.168.1.1 255.255.255.0

frame-relay interface-dlci 601

class shape_map_class



This archive was generated by hypermail 2.1.4 : Sat Sep 01 2007 - 11:32:11 ART