SCTP Discussions

Text archives Help


Re: [SCTP-Discussion] Peer restart question


Chronological Thread 
  • From: Michael Tuexen <Michael.Tuexen AT micmac.franken.de>
  • To: "buskila@gmail,com" <buskila AT gmail.com>
  • Cc: <discussion AT sctp.de>
  • Subject: Re: [SCTP-Discussion] Peer restart question
  • Date: Tue Aug 15 13:02:01 2006
  • List-id: SCTP Discussions <discussion.sctp.de>

Hi Oleg,

protocol question are best sent to tsvwg AT ietf.org. The discussion AT sctp.de list handles
implementation specific questions of the sctplib.

Answering your question: You need to inform the user at the B side that the A side
has not performed the graceful shutdown, but a restart. That is why the upper layer
gets a RESTART notification.

Best regards
Michael

On Aug 15, 2006, at 12:14 AM, buskila@gmail,com wrote:

Hello

I'm receiving the RESTART notification from the SCTP stack. The scenario is this:

0) Association is established between A and B
1) Side B calls sctp_shutdown, but the call is manually deferred right after the function returns (the association stated changes to SHUTDOWNSENT)
2) Side A crashes ungracefully
3) Side A restarts and initiates the association to B using the same source and destination ports
4) Side B continues and the ULP receives the RESTART notification


I was wondering if this is the correct behavior. The RFC states (in 5.2.4) that the restart rules apply in any state, however in this case it does not make much sense.

Any suggestions?

thanks
oleg






Archive powered by MHonArc 2.6.19+.

Top of page