DHCPv6, was: Re: IPv6 Finally gets off the ground

Mon Apr 16 22:13:10 UTC 2007

If you turn on IPv6 on an XP machine (or have it turned on for you
by a "helpful" application or MCP-enabled IT staff) be aware 
that there can be unexpected consequences.

In my case it was discovering the nooks and crannies of Teredo,
Microsoft's IPv6 tunnelling protocol.

http://en.wikipedia.org/wiki/Teredo_tunneling

I spent a couple hours in a hotel recently trying to untangle why
using the DSL system I could see the net but couldn't get to any
sites other than a few I tried at random like the BBC, Yahoo
and Google.

That's because they are among the few that apparently have 
IPv6 enabled web systems.

Once the reason became apparent, I found another terminal and
figured out how to disable Teredo and IPv6 on my laptop and all
was well for the duration.

Lesson learned.

I was once, circa 1995 or so, fairly enamored of IPv6.  Now it 
makes me wonder just exactly what problem it is good at solving.

Don't get me wrong -- it's not the fault of IPv6 and its designers
and advocates, it's that the world has moved on and other
methods have been found for the questions it was designed to 
address.  There is certainly room for concern about how well
those work, but the conversion effort to IPv6 -- well, the market
has voted with its pocketbook, or not.  Present company included.

fh

-----------------
>On Mon, Apr 16, 2007 at 01:59:36PM +1200, Perry Lorier wrote:
>> >When you can plug your computer in, and automatically (with no
>> >clicking) get an IPv6 address, 
>> 
>> Router Advertisements let you automatically configure as many IPv6 
>> addresses as you feel like.
>
>Remember that in XP, which Iljitsch recently cited to support his
>claim of "years of operating system support," you must click IPv6
>into your configuration.  It probably wants your XP install disc,
>or something like that.
>
>In my point of view, this does not cut the mustard for such words.
>
>
>Let's be clear:
>
>"There has been router and operating system support for years" is
>a statement which predicates that the World has no technical excuse
>for not running IPv6 globally edge-to-edge already.
>
>I think such a statement is fundamentally flawed.
>
>
>> This could be a fairly simple defacto standard if network operators 
>> start using it.  This is an obvious weak link in the chain at this point 
>> tho.
>
>Does this represent "years of router and operating system support?"
>
>My answer is "no."
>
>> once you have DNS you can use the WPAD proxy auto discovery thingamabob.
>
>...if you also had your domain suffix (unless you are suggesting
>that there have been WPAD records at the root for "years"?).
>
>RTADV won't help you here (tho they keep talking about putting
>domain-search and nameservers in it), and neither will DHCPv6
>as it turns out (it carries a domain-search list, but not "your
>domain suffix" which is more what WPAD should really want).
>
>This is not "years of operating system support."
>
>What has had "years of operating system support," is the
>unfortunate practice of acquiring option code 252 in DHCPv4.
>
>> >and solve your dynamic dns problems (as IPv4 set top boxes do today), 
>> 
>> Updating your forward/reverse dns via DNS Update messages isn't that 
>> uncommon today.
>
>On Enterprise networks using GSS-TSIG, sure.
>
>On ISP networks, I think the only time end-hosts try to update
>their reverse DNS directly is when they're participating in a
>rather unfortunate, and unintentional, distributed DoS against
>the root servers.
>
>Which, oddly enough, you mention next.
>
>Actual reverse dns updates for end hosts (and not their NAT
>gateways) is relatively uncommon, owing to the fact that such
>end hosts generally are on RFC1918 addresses.
>
>> http://www.caida.org/publications/presentations/ietf0112/dns.damage.html
>> 
>> where hosts are trying to update the root zone with their new names.
>
>I'm confused by what you're trying to argue.  Are you suggesting
>that AS112 represents "years of operating system support for
>IPv6"?
>
>> So you can get from A to D without requiring DHCPv6.
>
>...I hope you see that this is only so long as you require some
>clicking instead.
>
>This is all well and good for those of us who have sufficient
>growth (or equivalent feminine metaphor) on our chins, which we
>enjoy stroking thoughtfully while determining what all these
>"correct configurations" are.
>
>But I don't think "it works for bearded geeks" is setting the
>bar high enough when we use lofty words like "supported by
>routers and operating systems for years."
>
>-- 
>David W. Hankins	"If you don't do it right the first time,
>Software Engineer		you'll just have to do it again."
>Internet Systems Consortium, Inc.	-- Jack T. Hankins
>