RE: FRTS Theory Question

From: Michael (mamiller2@comcast.net)
Date: Wed Apr 05 2006 - 04:06:33 GMT-3


What is the difference of completing this task in policy-map vs. class-map
for FRTS? I know one is applied to service-policy. Any others? Does it
yield same goal? Any advantages or notes to one particular method?

Brians?

CR5-2651(config-pmap-c)#shape ?
  adaptive Enable Traffic Shaping adaptation to BECN
  average configure token bucket: CIR (bps) [Bc (bits) [Be (bits)]],
                  send out Bc only per interval
  fecn-adapt Enable Traffic Shaping reflection of FECN as BECN
  fr-voice-adapt Enable rate adjustment depending on voice presence
  max-buffers Set Maximum Buffer Limit
  peak configure token bucket: CIR (bps) [Bc (bits) [Be (bits)]],
                  send out Bc+Be per interval

-----------------------------------------------------------

CR5-2651(config)#map-class frame-relay ?
  WORD Static map class name

CR5-2651(config)#map-class frame-relay vc1
CR5-2651(config-map-class)#?
Static maps class configuration commands:
  default Set a command to its defaults
  exit-class Exit from static map class configuration mode
  frame-relay Configure Map parameters
  help Description of the interactive help system
  no Negate a command or set its defaults
  service-policy class-based service policy

CR5-2651(config-map-class)#fram
CR5-2651(config-map-class)#frame-relay ?
  adaptive-shaping Adaptive traffic rate adjustment, Default = none
  bc Committed burst size (Bc), Default = 56000 bits
  be Excess burst size (Be), Default = 0 bits
  cir Committed Information Rate (CIR), Default = 56000 bps
  congestion Congestion management parameters
  custom-queue-list VC custom queueing
  end-to-end Configure frame-relay end-to-end VC parameters
  fair-queue VC fair queueing
  fecn-adapt Enable Traffic Shaping reflection of FECN as BECN
  fragment fragmentation - Requires Frame Relay traffic-shaping to
be
                     configured at the interface level
  holdq Hold queue size for VC
  idle-timer Idle timeout for a SVC, Default = 120 sec
  interface-queue PVC interface queue parameters
  ip Assign a priority queue for RTP streams
  mincir Minimum acceptable CIR, Default = CIR/2 bps
  priority-group VC priority queueing
  tc Policing Measurement Interval (Tc)
  traffic-rate VC traffic rate
  voice voice options

-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of Petr
Lapukhov
Sent: Tuesday, April 04, 2006 10:33 PM
To: Todd.Osterberg@compucom.com
Cc: ccielab@groupstudy.com
Subject: Re: FRTS Theory Question

Hi Todd,

This question is being discussed over time on GS.

Basically, you lower Tc (smller burst) when you need
to deliver some sort of delay-sensitive traffic (voice).

The best thing you can read on this topic is
B. Dennis' presentation at

http://www.internetworkexpert.com/resources/01700368.htm

HTH
Petr

2006/4/5, Todd.Osterberg@compucom.com <Todd.Osterberg@compucom.com>:
>
> I am trying to understand how IOS works under the hood, so to speak,
> when performing frts. I understand the formula of T=Bc/CIR. If I
> configure a map-class with a CIR value why do I need to specify a Bc.
> Since the default of T is .125 seconds and I have then provided the CIR,
> why doesn't IOS solve for Bc? Here is my configuration example:
>
>
>
> Initial configuration:
>
>
>
> map-class frame-relay frts
>
> frame-relay cir 128000
>
>
>
> Rack1R1#sh traffic-shape
>
>
>
> Interface Se0/0
>
> Access Target Byte Sustain Excess Interval Increment
> Adapt
>
> VC List Rate Limit bits/int bits/int (ms) (bytes)
> Active
>
> 105 128000 2000 128000 0 125 2000 -
>
>
>
>
> When the map-class is modified to include a Bc, the results are
> different:
>
>
>
> map-class frame-relay frts
>
> frame-relay cir 128000
>
> frame-relay bc 16000
>
>
>
> Rack1R1#sh traffic-shape
>
>
>
> Interface Se0/0
>
> Access Target Byte Sustain Excess Interval Increment
> Adapt
>
> VC List Rate Limit bits/int bits/int (ms) (bytes)
> Active
>
> 105 128000 2000 16000 0 125 2000 -
>
>
>
>
> Any thoughts as to why FRTS works this way?
>
>
>
>
>
> Todd
>
>
>
>
>
>
>
> Todd Allen Osterberg
> Senior Consultant, Professional Services US Western Region
>
> CompuCom
> eMail: todd.osterberg@compucom.com
> T: (916) 577-1066
> C: (916) 316-4650
> www.CompuCom.com
> <outbind://4-0000000078027C4F809C5F448BBD4AD06CC1CD9207005BA671B4ADEFD54
> 18728E128462975250000000D51F50000B7CF08D69463BB4FA89400956DFA551A0000010
> 7CF9A0000/exchweb/bin/redir.asp?URL=http://www.CompuCom.com>
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html



This archive was generated by hypermail 2.1.4 : Mon May 01 2006 - 11:41:56 GMT-3