RE: Marker insertion question

The problem is that you don't really know that you have a complete FPDU;
you could have a data pattern that LOOKS like an FPDU, but is not, due
to a resegmentation that you were not expecting.  The group long ago
decided that the probability of mis-detecting a data pattern as an FPDU
in this case was not acceptable, so markers were implemented to answer
the challenge.
 
Paul R. Culley
HP Fellow
281-514-5543
 

________________________________

From: Ben Sum [mailto:[email protected]] 
Sent: Monday, August 28, 2006 8:41 AM
To: [email protected]
Subject: [rddp] Marker insertion question



If the FPDU is always ended within a TCP segment, is marker generation
really needed?  Even when the TCP segment is received out of order, the
receiver already received the complete FPDU and the marker value is
really insignificant.

 

Ben

 

 

Any content within this email is provided "AS IS" for informational
purposes only.  No contract will be formed between the parties by virtue
of this email.

_______________________________________________
rddp mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/rddp