Hi Tony,
Regarding Lixia's critique of Ivip, you wrote:
> Received and incorporated.
>
> I noted that there was no comment on the MTU issue. Should there be?
I am not sure what you are referring to.
Ivip is intended to run in the long-term, and ideally from the start,
with Modified Header Forwarding rather than encapsulation - thereby
avoiding the Path MTU Discovery problems which arise from encapsulation.
ETR Address Forwarding (EAF) - for IPv4
http://tools.ietf.org/html/draft-whittle-ivip4-etr-addr-forw-01
Prefix Label Forwarding (PLF) - for IPv6
http://www.firstpr.com.au/ip/ivip/PLF-for-IPv6/
Without these, I think Ivip would still have many advantages, so they
are not essential parts of Ivip. There has been only a little debate
about these.
They both require upgrades to all (or almost all) DFZ routers and
some other routers. In the long-term, this is not a problem since
new routers could be built and programmed to handle the new
functionality.
There are various things which could be debated, such as whether
these approaches would work and whether this would be a good use for
these header bits.
Also, there is the question of how much effort and expense would be
required to upgrade the routers, for instance, for IPv4
implementation later this decade - so as to avoid the need to do
encapsulation at all. Since I think there's less of a hurry for
IPv6, I figure there's a much greater chance of having the routers
ready for whenever Ivip is introduced for IPv6. Any router with a
firmware controlled FIB should be upgradeable without too much fuss.
There's space for someone to write a critique of this aspect of Ivip,
or to add to what Lixia wrote with any other concerns.
- Robin
_______________________________________________
rrg mailing list
rrg@irtf.org
http://www.irtf.org/mailman/listinfo/rrg
