Hi Wes,
On 10-10-25 02:37 PM, George, Wes E IV [NTK] wrote:
-----Original Message-----
From: Suresh Krishnan [mailto:suresh.krish...@ericsson.com]
Sent: Monday, October 25, 2010 1:57 PM
Subject: Re: I-D Action:draft-krishnan-6man-header-reserved-bits-00.txt
I strongly support the load-balancing use of the flow label. The
question is how many bits are really required. I am not an operator and
hence cannot speak authoritatively, but talking to our customers leads
me to believe that 16 bits is sufficient for this purpose. I would love
to see some more data on this.
[WES] I guess the question as it regards this draft is, what if the actual
answer is 17 or 18 bits? Is 2-3 bits still enough to reserve for an arbitrary
future use? For that matter, is 4? Why or why not?
Might be worth anonymizing and including some of what you've been told by your
customers as rationale for the decision for 16 bits in your draft, and then we
can debate whether that's accurate, or a good enough compromise because we
aren't as concerned about the random-guess attacks Brian references.
OK.
It is hard for me to guess all the future uses, but I will take one
probable use case: re-ECN.
[WES] ok, maybe - I'd have to do more research on that, both its implementation
and the likelihood of significant deployment before agreeing/disagreeing with
you. My point was: Document in draft along with hopefully other examples so
that we can decide if we care more about that than having another 4 bits for FL
loadbalancing :-)
Sounds good. We cannot get this done before the draft cutoff in 2 hrs
:-) . If you are going to be in Beijing, we can have a short discussion.
Cheers
Suresh
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
