The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2003-Jul> msg00089



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

RSVP Fastreroute

  • From: Gopalkrishna Panicker <gopanicker@yahoo.com>
  • Date: Thu, 17 Jul 2003 18:10:35 -0700 (PDT)

All,

I have a few questions on RSVP fastreroute (draft-ietf
-mpls-rsvp-lsp-fastreroute-03.txt)

Consider the following network,

A---B---C---D---E---G
    |   |       |
    |---F-------|

Assume the following
The protected LSP has a ERO (ABCDEG)
1-to-1 backup is being employed.
All nodes are using Sender Template Method of backup
identification

Also,
B creats a backup LSP with ERO (BFEG)
C creats a backup LSP with ERO  (CFEG).

If I understand section 7.1.1 (Merging Backup Paths
using the sender template specific method) correctly,
one (and only one) path message will be sent/refreshed
from nodes F to E. The sender template in this path
message will belong to the winner of this merge. 
Is this correct ? If so shouldn't the traffic spec of 
this final path message reflect the traffic
requirements of the messages that have been merged.

Is this what the authors mean when they say in 
paragraph 2 (line # 2)of section 7.1.1,
 "Similar to that specified in [RSVP-TE] for merging
of RESV messages,". 

So, if fastreroute is not being used, merging
will still occur the way I have described above
(assuming ofcourse that I am correct in the first
 place).  If not, at a detour merge point how can 
I tell the difference  between a detour LSP ( no
 DETOUR object  ) and a regular non-fastrerouteable
LSP i.e one should be merged and one should not.
  
Thanks in advance,

Cheers,
Gopal
  


__________________________________
Do you Yahoo!?
SBC Yahoo! DSL - Now only $29.95 per month!
http://sbc.yahoo.com


  • Follow-Ups: