From: Christian Huitema (huitema_at_no.spam)
Date: Mon Mar 25 2002 - 09:01:51 PST
This is the second response to Pekka's comments, addressing his concerns
about the different types of NAT:
> - I think we also need a section on what happens if the assumptions
> about NAT's don't hold about Teredo-Teredo shortcuts (like 4.1.4)
>
> - Has the behaviour described in 4.1.4 been tested in practise? I'm
> mostly curious about the more restrictive (port restrict, symmetric)
> cases. I'm having hard time visualizing how the bubbles' ports are
being
> mapped in different NAT's.
>
> ...
>
> - I'm very worried about different NAT types.. I'd really like to get
> more concrete data on the behaviour.
The basic answers are: yes, we have tested the behavior described in
4.1.4, and it works in practice in very many cases. When the NAT don't
quite behave in the expected way, the behavior falls back to simply
relaying all the packets through the servers.
-- Christian Huitema
This archive was generated by hypermail 2.1.7 : Fri Oct 06 2006 - 00:00:32 PDT