Hi Joe, > -----Original Message----- > From: Joe Touch [mailto:[email protected]] > Sent: Friday, February 27, 2015 3:35 PM > To: Templin, Fred L; Ronald Bonica; Carlos Pignataro (cpignata) > Cc: [email protected] > Subject: Re: [Int-area] draft-ietf-intarea-gre-ipv6 > > > > On 2/27/2015 3:22 PM, Templin, Fred L wrote: > > Hi Joe, > ... > >> I.e., this is consistent with IPv4 allowing on-path fragmentation, and > >> consistent with IPv6 never fragmenting on-path. > > > > You are being dogmatic for no good reason. > > RFC2460 forbids it *explicitly*. > > That's a good reason.
You are still being too dogmatic. For atomic fragments, the situation is no different than for an IPv6/IPv4 translator. In fact, it is even better because the IPv6 tunnel ingress knows that the final destination is required to reassemble at least 1500. For an IPv6/IPv4 translator on the other hand, all the translator is assured is 576. I do not see any "MUST NOTs" in RFC2460, nor any dire consequences for allowing a tunnel ingress to fragment atomic fragments. So, if we update the final paragraph of Section 5 of RFC2460 we should be good to go. Thanks - Fred [email protected] > Joe _______________________________________________ Int-area mailing list [email protected] https://www.ietf.org/mailman/listinfo/int-area
