The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2002-Feb> msg00097



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

Doubt about Refresh reduction

  • From: Apratim Mukherjee <ApratimM@netbrahma.com>
  • Date: Wed, 13 Feb 2002 16:19:48 +0530

Hi ,

Doubt #1 :
Lets take a set-up
		 |-----------------N6---------------------|
		 |    			        |
	------N1 ------ N2 ----- N3 -----N4 -----N5-----

	N1 is sending summary refresh to N2 for a Path message destined to
N5 currently going thru n2 , n3 , n4 

	so , if  a route change occurs for only this path ( link betwn n4
and n5 goes down) then should it not send a trigger refresh immediately to
N6 and stop sending summary refresh to N2 ? 
	If , in a specific implementation , there is no route change
notification , then should it check for each path refresh it sends out in
form of msg id, whether route has changed so that the path message should
now go thru another next hop ?

	Doubt # 2 :
>From RFC2961 abt refresh reduction:
"Srefresh NACKs are used to indicate that a received
Message_Identifier field carried in MESSAGE_ID LIST, SRC_LIST, or
MCAST_LIST object does not match any installed state.  This may occur
for a number of reasons including, for example, a route change."

Could someone tell me how this would occur ?
If my previous assumptions are correct then the previous hop will not send a
refresh summary message for a path message whose route has changed . i feel
that only if a path tear or resv tear has happened after a route change then
at node N1 if a resv refresh msgid is sent from N2 after route change has
occured and then a state teardown , will this sort of a situation arise
.Please correct me if i am wrong .

Thanks 
Apratim