Hi Thomas, I guess Bert's comparison comes from the fact we can decide the path a packet takes, just like we can for MPLS. So if the Path Computation Element(PCE) calculated a path using CSPF that had loops(and ence signalled using RSVP-TE), a similar situation could arise.
However I disagree with Bert here. The difference is that MPLS is connection oriented in nature. The path setup element(RSVP-TE or management for T-MPLS) and the data packets are different flow. It would require to at least hack the head end router to get a simialr functionality. Besides off-path attacks will be harder to make. Thanks, Vishwas On 8/30/07, Thomas Narten <[EMAIL PROTECTED]> wrote: > > As always, there are workarounds like MPLS, if RH0 is eliminated. > > I don't get this at all. RH is end to end (i.e., it's part of the IP > header, and presumably added by the originating device and processed > by the ultimate destination). MPLS, on the other hand, is an L2 > technology and does not involve either the original sender or the > ultimate destination (normally). > > I see them providing completely different services and NOT being > interchangable in any real sense. > > Thomas > > -------------------------------------------------------------------- > IETF IPv6 working group mailing list > ipv6@ietf.org > Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 > -------------------------------------------------------------------- > -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------