Re: Fwd: [ipv6mib] So, where were we?

["C. M. Heard" <heard@pobox.com>]

On Fri, 4 Oct 2002, Margaret Wasserman wrote:
> There have been some recent discussions with the IPv6 MIB design
> team regarding the direction of our work on IPv6 MIBs that I
> think would be best discussed on the full IPv6 and MIBs lists.
> 
> In particular, we have received input from two areas:
> 
> 	(1) Folks who would like to see substantial changes
> 		to the RFC 2096 update (Forwarding Table MIB),
> 		to make it better reflect the structure of
> 		forwarding tables on current equipment, be more
> 		efficient to use, etc.

I have implemented RFC 2096 once on a very simple end system.
While I concede that it was a big improvement over the broken
ipRouteTable in RFC 1213, I could not help but think while
writing the code how difficult it would have been to implement
efficiently in any kind of complex equipment.

Based on that experience, I tend to believe that a major
restructuring is probably necessary if you want something
that is both logically consistent and reasonably efficient
to implement.  I don't think that a simple restructuring
of RFC 2096 to be IP version-neutral will gather much
support among implementors.

At the very least, I would urge that you ask for deployment
reports for RFC 2096 before deciding that a simple
version-neutral update is the right way to go.

Mike