The MPLS-OPS Archive

["alok" <alok.dube@apara.com>]

hi

 

is the discussion based on RFC 3346 or is the discussion based on "why MPLS"??

 

i mean as far as i see that RFC it simply says "MPLS can achieve TE" ...

 

MPLS alone has its own advantages im sure.... but im not able to gather if people here are asking the validity of having an new encapsulation, the validity of TE or what?

 

I read that Jennifer Rexford's  too, fair enuf, given those set of condtions that "all is IP",i want to do is TE, no MiBGP etc....its fine for that model..

 

I think the point is that "MPLS does TE" , so do intserv models like RSVP-TE with IGPs etc...

 

...and a well modelled diffserv network too wud do it...if u dont oversubscibe bandwdiths i guess...

 

but there are other areas where u might need MPLS or some such label protocol...

 

I dont quite gather the gist of what we are talking about here....

 

 

 

----- Original Message -----

From: Islam, Aziz

To: mpls-ops@mplsrc.com

Cc: Mathew Lodge ; alok ; M. ELK

Sent: Wednesday, October 30, 2002 11:45 PM

Subject: RE: [MPLS-OPS]: IGP tuning for TE versus MPLS-TE

Folks,

Another well-written article can be found in the October issue of IEEE Communications magazine. The title is "Traffic Engineering with Traditional IP Routing Protocols".

 

Aziz S. Islam
Sr.Infrastructure Splst.- CCIE(R/S)
EDS-Design Engineering
33 Yonge Street; Suite 400
Toronto, Ontario M5E 1G4
CANADA
phone: (416)814-1696
pager: (416)517-4985
mailto:aziz.islam@eds.com
http://www.eds.com

-----Original Message-----
From: Mathew Lodge [mailto:Mathew@CPlane.com]
Sent: October 30, 2002 11:25 AM
To: alok; M. ELK
Cc: mpls-ops@mplsrc.com
Subject: Re: [MPLS-OPS]: IGP tuning for TE versus MPLS-TE

All,

There is an excellent discussion of optimal routing done via IGP metrics (no MPLS) vs. explicit path selection (MPLS TE) from AT&T. You can find it at

http://www.research.att.com/~jrex/papers/ieeecomm02.long.ps
http://www.research.att.com/~jrex/papers/ieeecomm02.long.pdf

Note that this paper is a discussion of theory -- i.e., it addresses how close can you get to optimal routing using Dijkstra's SPF algorithm with judicious link weight selection, and what happens to the efficiency of the network when links fail and IGP re-routes traffic. In practice, I wonder how disruptive it would be to a regular IP network to go around tinkering with the link weights on every router with an offline TE tool (which this approach mandates).

Cheers,

Mathew


At 03:26 PM 10/30/2002 +0530, alok wrote:

what do u do for "non ip" traffic ;o)
 
can u "TE" it too with IGP tunning?
 
 
 
-rgds
alok

----- Original Message -----
From: M. ELK
To: mpls-ops@mplsrc.com
Sent: Wednesday, October 30, 2002 3:01 PM
Subject: [MPLS-OPS]: IGP tuning for TE versus MPLS-TE

Their is some hot debate previously on this list about an article on MPLS which stated that no need for MPLS ,anyhow the conclusion from this debate was that the writer of such article is clueless .
here is a doc from sprint which state exactly the same :
http://www.sprintbiz.com/business/resources/resource/SprintCiscoMPLS.pdf
(guess ,could not state that they are clueless too . also the doc have the cisco logo ) .
 
need not to start another hot debate ,just need U comment on the possibility
and limitation of TE by tuning IGP versus MPLS-TE , any other provider
droped MPLS-TE favoring IGP tunning ? .
 
Brgds

---

Get a speedy connection with MSN Broadband.  Join now! Click Here ------- The MPLS-OPS Mailing List Subscribe/Unsubscribe: http://www.mplsrc.com/mplsops.shtml Archive: http://www.mplsrc.com/mpls-ops_archive.shtml

| Mathew Lodge                 | mathew@cplane.com     |
| Director, Product Management | Ph: +1 408 789 4068   |
| CPlane, Inc.                 | http://www.cplane.com |