Hi all,
After looking at the responses on the mailing list, we find that
there is consensus to change MAP-T into a Standards Track document.
However, there was an issue raised regarding preservation of DF=1 and
MF=1 and this warranted inclusion of some advisory text in the draft.
The following text has been added (Thanks Rémi and Tom) in section 10 of
version -07.
"Note: The NAT64 [RFC6145] mechanism is not entirely losseless and when
applied to IPv4 communication paths that traverse double NAT64 networks,
IPv4 originated ICMP-independent PathMTU Discovery as specified in
[RFC4821], ceases to be entirely reliable. This is because the[RFC4821]
defined DF=1/MF=1 combination, after NAT64 translation, becomes
DF=0/MF=1."
Authors, Please resubmit a new version (-08) on Standards Track.
Thanks
Suresh & Yong
_______________________________________________
Softwires mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/softwires
