CCing the IAB because I think we are reaching a slippery architectural
slope. Hopefully they can help us out.
On 7 aug 2009, at 20:43, Shane Amante wrote:
Therefore, I'll have to revise my original recommendation in the
first bullet above that we only consider UDP with 0 checksums as
the preferred short-term solution when IPv6 is being used as the
outer encapsulation,
I don't see that. Currently, there isn't that much IPv6 traffic in
the first place, and certainly not between the same source/
destination addresses. So the lack of a fine-grained optimal
solution to the load balancing issue is not a problem in practice.
This affords us the relative luxury of being able to ignore current
problems in implementations and do the right thing, rather than be
forced to do something ugly and difficult (that would be UDP with
no checksum).
In that case, I guess we'll have to agree to disagree
Well, we still either have UDP encapsulation with a 0 checksum for
LISP over IPv6 or not.
and in doing so, I'd like to take the liberty of quoting Vince
Fuller's message to this list back on 8/4/2009, where he says it
better than I could:
---snip---
> Current operational reality is that the installed base of transit
routers
> on the Internet uses a hash of source/dest addres/port to split
traffic
> across LAGs so LISP uses an encapsulation that is compatible with
that
> reality.
>
> Specifying some alternate reality and hoping that the operational
world will
> modify its behavior to match doesn't seem very practical,
particularly since
> one of LISP's virtues is that it requires no changes to the
transit routing
> system.
---snip---
The problem with adopting that logic and adopting a UDP header for
LISP even though that would require breaking the IPv6 spec just
because router vendors haven't gotten around to implementing ECMP
based on the flow label is that it cements us into a TCP/UDP view of
the world with IPv6 just like NATs did for IPv4.
In my opinion, the possible short term pain of less than optimal load
balancing in a few places with high IPv6 LISP traffic is worth it if
that allows us to proceed in an architecturally clean way.
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------