That helps, but understanding of IPv6 and mindshare is even harder than
I'll agree that it is hard. That's why the clue x 4 keeps having
to be applied.
That's a LOT of people to whack on the side of the head. And it pretty
much has to be done in meatspace; the net doesn't help you out with this
And if you start
looking for technology that would let you automate renumbering your
entire network, you might find that the technology that exists is
incomplete and unproven.
Which is why I keep saying. Run through the renumbering exercise.
Find the problems. Report them to your vendors. Vendors being
proactive would be a big help here.
Yes they would. But basically everyone can assume that this is Somebody
Else's Problem. You want to make it the problem for the network admins,
sysadmins, users, application writers, and firewall vendors to solve.
Those people want to make it the problem for the carrier ISPs and
router vendors to solve.
I've spent a lot of time adding support to make renumbering
easier in the places that I can change. I will continue
to do that.
I can't however fix every problem.
And really, fixing the routing system so that it can provide stable
global PI addresses to everyone (say, via something LISP-like) might be
easier...especially that the need to renumber isn't the only problem
that lack of stable global PI addresses causes.
oh yes, and practical use of DNS security still seems to
It will as long as people don't actually sign there zones.
Have you asked for cs.utk.edu to be signed?
I don't work there any more, so it's Somebody Else's Problem. :)
And really, there's no way I'd trust DNS to do this. I've spent too
many years watching it break.
It breaks much less often once you allow it to be automated.
We could in theory have all nameservers update the parent
servers with appropriate glue. This is not technically hard
to do. Similarly NS RRsets.
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742 INTERNET: Mark_Andrews(_at_)isc(_dot_)org
Ietf mailing list