Re: [TE-wg] TE use in today's networks

[Yakov Rekhter <yakov@juniper.net>]

Noel,

>     >> I think, manual "Optimization" of weights (even small number if
>     >> weights) based on the huge topology is a real disadvantage.
>  
>     > this doesn't have to be done manually -- it could be automated. The
>     > basic idea is to know the current topology and offered load in your
>     > network by measurement and to have a tool that can select weights that
>     > satisfy some objective function (and, perhaps, try to minimize the
>     > number of weights that need to change from the existing configuration).
>     > Then, you need a way to effect weight changes in the network
>  
> Ahem. I feel a major, major rant coming on.

Attached e-mail that was posted to the irtf-rr mailing list seems to be
relevant to the discussion.

Yakov.
----------------------------------------------------------
From: Yuval Shavitt (shavitt@eng.tau.ac.il)
Date: Thu May 17 2001 - 08:41:54 EDT 


In the paper below we give an example that shows that regardless
of the weights you assign to links in OSPF the possible gap between
MPLS TE capabilities and OSPF which equally splits the traffic
among multiple paths in as high as can be.

In other words, weight setting for OSPF cannot replace MPLS as 
a traffic engineering tool. 

Yuval 

http://www.eng.tau.ac.il/~shavitt/pub/DIMACS01-17.ps 

> 
> Our work is basically a theoretical result with an attemp to find some 
> practical applications. 
> I agree that it is too early to say how ready or how ususful it can be for 
> real world networks. 
> I am aware of some of its limitations. One of them is that it sometimes 
> needs to split traffic 
> unequally between two or more equal weight shortest paths in order to 
> achieve the optimal 
> traffic engineering which is doable by MPLS (uneqal split between two LSPs 
> between the same 
> source-destination) but for some reason difficult to do by OSPF at present 
> time. Another is that 
> we may be able to set appropriate link weights to optimize the long term 
> network performance in the 
> statistical sense, but we don't want to change link weights dynamically 
> reacting to traffic 
> fluctuations in real time. I welcome critics or sugegstions to make it more 
> practical. 
> 


-- 
Yuval Shavitt
Dept. of Electrical Engineering - Systems
Tel Aviv University, Tel Aviv 69978, Israel
Tel: +972 3 640 8659  Fax: +972 3 640 7095
URL: http://www.eng.tau.ac.il/~shavitt
Networking Research Center, 
Bell Labs, Lucent Technologies
Holmdel, NJ 07733
URL: http://www.cs.bell-labs.com/~shavitt