Brian E Carpenter wrote:
On 2007-07-03 11:27, Perry Lorier wrote:
And actually that is where we seem to be at odds. I fear that ULA-G
will leak.
That's all.
So I might be a bit naive here, but why is this a problem? ISP's
that are worried that they're routers can't deal with ULA addresses
in their DFZ will filter all ULA addresses. ISP's that want ULA's in
their DFZ will make sure they dimension their routers such that they
can deal with it. Other ISP's will find some kind of middle ground
where they have some ULA prefixes,The DFZ is certainly not the same
everywhere on the Internet, by making ULA's easily filterable it
gives operators a place to start if they want to filter.
Of course. But by leak, as I think was clear in later mail, I meant leak
deliberately under economic pressure. Accidental leakage would be
statistically insignificant, I assume, and easily filtered out as you
say.
Brian,
Can you outline the economic pressures that you believe would lead to a
ULA-C/ULA-G being accepted by more the majority of tier1ish transit
providers in the DFZ? Like Paul, I'm inclined to believe the
countervailing economic pressures would be much stronger, and that we
will see very few, if any, ULA routes crossing the magic threshold of
global routability. I would love to hear and possibly debate the
counterarguments.
Thanks,
Scott
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------