Re: Flow label versus Extension header - protocol itself

[Margaret Wasserman <margaret@thingmagic.com>]

Hi All,

At 12:14 PM +1000 5/18/05, Geoff Huston wrote:
> > => I am sorry but I believe to propose major changes in IPv6 is no more
> > a good idea today. Perhaps you have a little deployment where you are
> > but don't expect it is the case everywhere...
>
> I suspect that Shim6 is more of the nature of a "major change" than
> an "insignificant tweak", and the concept of explicit use of extension
> headers  for signalling appeals to me much more than hacking away at
> header bits in the flow label. I would suggest that we should indeed
> attempt to engineer this using extension headers where extension
> headers are logically required.

I agree with Geoff -- SHIM6 will represent a major change to IPv6, 
anyway, so adding an extension header is not unreasonable.

My concern (once upon a time) with extension headers is that the use 
of extension headers on some packets and not others effectively 
presents a variable MTU to upper layers.  We need to make sure that 
this is properly handled to avoid bad interactions with PMTU 
discovery code.

Margaret