RE: Evolution of the IP model - ICMP and MTUs

[Christian Huitema <huitema@windows.microsoft.com>]

If you want to document the evolution, you have to be complete.

1) In the original model, senders of datagrams with the DF bit set
 (Don't Fragment) received  no information back.
2) In 1990, the Next-Hop MTU information was added to Datagram Too Big
 ICMP message (RFC 1191).
        Hosts have a chance to discover the real MTU in the path using ICMP
3) Around 1995, firewalls started to drop all ICMP by default
        Hosts that rely on ICMP to discover PMTU observe terrible performance
4) Around 2000, broadband connections start being equipped with tiny "home
   routers" whose NAT function does a pretty bad job at reassembling IP packets
        Hosts that send packets too large observe terrible performance, and they
        are in a bind since PMTU discovery does not work well.
5) By 2008, the IETF might recognize that firewalls are here to stay,
   that we could just as well forget about ICMP, but that we really
   need another solution.




> -----Original Message-----
> From: owner-v6ops@ops.ietf.org [mailto:owner-v6ops@ops.ietf.org] On
> Behalf Of Rémi Després
> Sent: Thursday, August 14, 2008 11:40 PM
> To: Dave Thaler; v6ops
> Subject: Evolution of the IP model - ICMP and MTUs
>
> Dave,
>
> I find your work on the IP model quite interesting.
> To answer your invitation to send comments, one remark.
>
> It concerns Path Maximum Transmission Unit Discovery (PMTUD), and more
> precisely its relationship to ICMP and Fragmentation.
>
> - In the original model, senders of datagrams with the DF bit set
> (Don't
> Fragment) received  no information back, when these packets were
> discarded on their way because of their being "too big", about what
> size
> would have fit.
> - In 1990, the Next-Hop MTU information was added to Datagram Too Big
> ICMP message (RFC 1191).
> - Thus, hosts have a chance to quickly discover exact path MTUs. (They
> have only "a chance" because: (1) ICMP messages, like any other
> packets,
> are returned in best effort mode; (2) some routers may have not
> implemented Next-Hop MTUs; (3) some firewalls filter ICMP packets.)
> - The upper layer TCPs can now take advantage of this IP-layer provided
> information to adjust their  Maximum Segment Size MSS (sec. 3.1 of RFC
> 1191).
>
> IMHO, this evolution would be worth noting in a next version of your
> draft.
>
>
> Regards,
>
> Rémi
>
>
>
>