On Sep 14, 2007, at 2:22 AM, David Conrad wrote:
And I would suggest by ignoring history we are doomed to repeat
it. I am not engaging in "I told you so" because I didn't --
you'll note I used "we". I am merely pointing out that we're
either at or very quickly approaching a crossroads and the choices
we have are constrained by the reality of the Internet today and
past decisions we, the IETF, have made.
Well, yes. But I do find myself wondering what tool one might really
want to use here and how it differs from what we do in IPv4.
Correct me if I am wrong (but not here - let's have that discussion
on v6ops). To my way of thinking, the process described in RFC 4192
can't really be automated start to finish, and it is nonetheless
pretty much the right process. Parts of it can be, such as once an
operator decides he wants to add a new prefix to every router
interface in his network, the database he uses to manage such things
can ssh to each router and add the prefixes, and similarly when he
decides to later remove the old, the database can do that. But the
big problem in renumbering isn't "getting the addresses assigned". It
is "finding and fixing all the places where that address was used in
numeric form to ensure that they now have the right new value". Since
human screwup behavior isn't automated, fixing human screwups is
difficult to automate.
So we can have tools that help with the major steps, but a lot of the
verification process can only be done by observation.
Recriminations and rants aren't going to make that much different.
What would be Really Nice would be to in some way ensure that
applications never saw IP addresses at all - they *only* worked on
names, and maintained no knowledge in the application of what address
was used. To my small mind, forcing a new DNS lookup in the event of
a TCP session failure and restart would be a good thing. The authors
of RFC 4778 would take exception - they want to be able to log into
the right place when everything is in flames. Apart from that,
though, managing addresses through names would go a *long* way toward
making renumbering easier. We already have many of those
capabilities, though. We have to as an industry consistently use them
that way.
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf