Re: [BEHAVE] NAT64 for Mobile IPv6 - Extension header translation ICMP error issue

> >Dan Wing wrote:
> >By the way, I see that both draft-sarikaya-behave-mext-nat64-dsmip-00
> and
> >draft-haddad-mext-nat64-mobility-harmful-01 state that IPv6  extension
> >headers are not translated and will cause the NAT64 to  generate an
> ICMP
> >error.  However, according to  draft-ietf-behave-v6v4-xlate an ICMP
> error is
> >not generated in that  case - draft-ietf-behave-v6v4-xlate says the
> extension
> >header is  simply skipped and the IPv6 packet is translated to an IPv4
> >packet.
> 
> 
> In order to start route optimization, MN sends Home Test Init (HoTI)
> message to
> IPv4-only server by tunneling it to HA. HA removes the tunnel header
> and the resulting packet goes through NAT64 packet translation. As Dan
> says
> above, the extension headers are simply skipped but this packet has no
> payload,
> I think that is the reason for HA generating an ICMP error and sending
> it to MN.

Oops, I had missed that "NAT64" in draft-haddad-mext-nat64-mobility-harmful
and in draft-haddad-mext-nat64-mobility-harmful refers to stateful
translation.  I had thought it was referring to stateless translation.

If it's a stateful NAT64, and the HoTI isn't run over UDP or TCP,
the stateful NAT64 will generate an error.  Text from 
draft-ietf-behave-v6v4-xlate-stateful-11:

   If the incoming packet is an IPv6 packet that contains a protocol
   other than TCP, UDP or ICMPv6 in the last Next Header, then the
   packet SHOULD be discarded and, if the security policy permits, the
   NAT64 SHOULD send an ICMPv6 Destination Unreachable error message
   with Code 3 (Destination Unreachable) to the source address of the
   received packet.

My mistake.  Sorry about that.

-d

> 
> Regards,
> 
> Behcet
> 
> 
> ----- Original Message ----
> > From: Behcet Sarikaya <behcetsarikaya@yahoo.com>
> > To: behave@ietf.org
> > Cc: mext@ietf.org
> > Sent: Tue, July 6, 2010 3:34:22 PM
> > Subject: [BEHAVE] NAT64 for Mobile IPv6
> >
> > Hi,
> >   We submitted a draft entitled NAT64 for Dual Stack Mobile IPv6
> which is
> at:
> > http://tools.ietf.org/id/draft-sarikaya-behave-mext-nat64-dsmip-
> 00.txt
> >
> > This  draft develops a solution to the problems described in
> >draft-haddad-mext-nat64-mobility-harmful by integrating NAT64
> operation with the
> >home agent and binding cache.
> >
> > We go one step ahead of NAT64 and provide  multicast support as well.
> >
> > Regards,
> >
> > Behcet
> >
> >
> >
> > _______________________________________________
> > Behave  mailing list
> > Behave@ietf.org
> > https://www.ietf.org/mailman/listinfo/behave
> >
> 
> 
> 