The MPLS-OPS Archive

Cell Relay Retreat>MPLS-OPS Archive>month:2001-Mar> msg00167



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

About RSVP-TE scalability - please help

  • From: Srihari Raghavan <sraghava@vt.edu>
  • Date: Mon, 26 Mar 2001 14:47:24 -0500
  • Importance: Normal
  • Resent-Date: Mon, 26 Mar 2001 16:40:13 -0500
  • To: mpls-ops@mplsrc.com

Hi all,
   I have some questions regarding RSVP/RSVP-TE. Can anyone please help me
with it?

   Am I right in saying the following things about RSVP/RSVP-TE when used
for MPLS traffic engineering at the core/backbone networks?

   1. RSVP in its original form is used to set up per-flow (data traffic)
state at each intermediate router. Is this found not scaleable because of
the 'soft state', potential large number of data flows and non usage of
reliable protocols like TCP?

   2. RSVP-TE, by contrast, is ONLY used to set up CRLSPs or ERLSPs and not
'flows' like that as MPLS 'traffic trunks'. Because, the number of LSPs (CR
or ER) can be far less than the number of traffic flows through it, RSVP-TE
is more scalable?
      Or, is it the case that the 'labels' being distributed by the egress
using RSVP-TE defines the traffic across the LSP and since 'labels' and LSPs
can be subject to label merging and LSP aggregation, RSVP-TE is scalable in
the core? Or is there anything that I am missing here about RSVP-TE
scalability?

   I think what i am trying to understand is how RSVP-TE is scalable while
there are some concerns about RSVP not being so. Is it because of
scalability extensions like like message_id extension, Bundle message
extension, summary refresh extension and hello protocol extension? or is it
any of the reasons above?

   3. Running diffserv with MPLS, am I right in saying that ERLSPs or CRLSPs
are set up by RSVP-TE with preset labels choosing different levels of QoS
guaranteed LSPs and diffserv running at Ingress, decides which LSP to choose
by appending labels to the incoming packets based on their DSCPs?


Can any one please confirm the above ideas about RSVP-TE, as its
understanding is very important to my work.

Thank you for your time

Srihari Raghavan
Graduate student
Dept. of Computer Science
Virginia Tech
=========================

-------
The MPLS-OPS Mailing List
Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
Archive: http://www.mplsrc.com/mpls-ops_archive.shtml