André, I agree. Added some text to 5.2.2
Bruce On Thu, Feb 2, 2012 at 5:38 PM, André Becker <[email protected]> wrote: > Hi, > > I'm currently implementing parts of RELOAD as my bachelor thesis and > stumbled upon an ambiguity concerning answers to fragmented requests. > Section 5.2.2 states: > > When a peer sends a response to a request using this routing > algorithm, it MUST construct the destination list by reversing the > order of the entries on the via list. This has the result that the > response traverses the same peers as the request traversed, except in > reverse order (symmetric routing). > > However, neither this section nor section 5.7 (which covers fragmentation) > clarifies which via list should be used if the message was fragmented. As > intermediate nodes handle each fragment independently, different fragments > of the same request might take different routes in the overlay. So for the > receiving node it remains unclear which via list should be used for the > answer. > > Proposal: Add the following text to section 5.2.2: > > If the request was fragmented and for that had to be reassembled by the > receiving node, the via lists of the fragments may differ. The receiving > Node > SHOULD use the via list of the last fragment, as determined by the last > fragment bit; See Section 5.7 > > This should be a good approach, because for the last fragment's route the > probability is high to still be available. > > Regards, > André > _______________________________________________ > P2PSIP mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/p2psip _______________________________________________ P2PSIP mailing list [email protected] https://www.ietf.org/mailman/listinfo/p2psip
