[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: ocean: do not boil
- To: email@example.com
- Subject: Re: ocean: do not boil
- From: Rob Austein <firstname.lastname@example.org>
- Date: Mon, 23 Sep 2002 18:07:50 -0400
- Delivery-date: Mon, 23 Sep 2002 15:08:08 -0700
- Envelope-to: email@example.com
- User-agent: Wanderlust/2.8.1 (Something) SEMI/1.14.4 (Hosorogi) FLIM/1.14.4(Kashiharajingū-mae) APEL/10.3 Emacs/20.7 (i386--freebsd) MULE/4.0(HANANOEN)
At Mon, 23 Sep 2002 23:41:46 +0200, Hesham Soliman (EAB) wrote:
> But why would they scale well for a large v4 cloud if v4 NATs are
Hah, I knew someone would spot that.
Because you don't use 'em for a large v4 cloud. You use 'em for a
little v4 edge site behind some kind of combination NAPT/NAT-PT box
and use some kind of v6-over-v4 tunneling over the rest of the v4
cloud. So you've got a lot of little translators converting to IPv6
as early as possible, which scales much better than big honking
translators at the border between the cloud and the rest of the world.
The same approach would "work" for large v6 clouds, but the result
would not be useful, because it'd require so many IPv4 addresses that
there'd be no point to the exercise.
["How do you climb down off an elephant? You don't. You climb down
off a duck, because it's so much safer."]