(sorry to spam 3 wgs, i have not received response to subscription
request for [EMAIL PROTECTED] yet)
> draft-ata-ipv6-anycast-resolving-02.txt
> A Protocol for Anycast Address Resolving
due to the way the proposal resolves (or hides) anycast address
into unicast address, we will experience a severe problem. application
thinks that it is communicating with anycast address AA, but ARL
(basically a NAT within client) translates it into unicast address.
therefore, applications which embeds address into its protocol payload
(like ftp) won't work as expected. section 4 (applicability statement)
is not true.
appendix A (how to map anycast address into unicast) basically has no
security considerations. section 5 (security consideration) is also
too weak.
another issue with this approach is, that anycast is being used as
service discovery mechnaism on the first contact only. anycast has
other benefits such as failure recovery/tolerance. these benefits are
gone with this draft.
itojun
--------------------------------------------------------------------
IETF IPv6 working group mailing list
[EMAIL PROTECTED]
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
