Re: (ngtrans) Re: IPv6 dns

[Robert Elz <kre@munnari.OZ.AU>]

    Date:        Sat, 20 Jan 2001 21:13:22 +0900
    From:        itojun@iijlab.net
    Message-ID:  <23571.979992802@coconut.itojun.org>

  | 	could you give us more concrete example, so that we can share the
  | 	image in your mind?
  | 
  | 	(i am 100% guessing here) if you mean something like dialup home
  | 	networks, wasn't it our goal to assign static address for these
  | 	networks?

No, not that specifically, though including that.

The point was that most systems on the net (that have addresses,
and are in the DNS) never get looked up, they just exist there.
A small fraction of everything that exists is the target of almost
all of the queries.

Yet, when there is a net renumbering, they all get renumbered, not
just the subset that the world cares about.   That tends to bias
the numbers somewhat away from the "millions of lookups per renumber
event" that was implied.

As for static addresses for dialup home nets - I'm not sure that I
care a lot, I'm not sure "dialup" nets will actually exist in significant
enough numbers for it to matter one way or the other very far into
the future.   Or at least I hope they don't.

And in the same thread ...

Antonio Querubin <tony@lava.net> asked:
  | Now, how many times would the same single workstation access some
  | shell server, or web server, or mail server, or whatever, somewhere
  | out on the net during that year?  How about in just one day? 

Never - they're all filtered.   But they're all still in the DNS.
And they do all get renumbered from time to time.  I doubt that
we're alone in having large collections of firewalled local hosts.

And to revisit Itojun's point - yes the goal is to have even all of
those have global (static, between renumberings) IPv6 addresses.
(And we happen to have been on the net long enough that we are lucky
enough to have sufficient global IPv4 addresses that they all have
global addresses now).

kre