[arin-ppml] ULA-C and reverse DNS

[Owen DeLong]

On Mar 22, 2010, at 12:15 PM, Michael Richardson wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> 
>>>>>> "michael" == michael dillon <michael.dillon at bt.com> writes:
>>> I think that it makes far more sense to make a liberal GUA policy
>>> that allows people to get GUA if they need it regardless of
>>> whether they need it for internet or not.  Then, if they want it
>>> from a prefix set aside as "non-routable", then, that's
>>> available, but, it's a purely advisory semantic, not something
>>> coded into systems or routers or whatever.
> 
>    michael> That is as bad as PA addressing. Your address range is
>    michael> tainted as unroutable, and if you want to change that, you
>    michael> have to return the addresses and get a new range and
>    michael> renumber.
> 
> I think that this is something to repeat multiple times.
> 
> I think that a lot of people regard address space as so valuable that it
> would be crazy to "waste" it by having two IPs addresses on a single
> machine.
> 
> This is where I think the notion that people will pay $$$$ to have their
> address space routed by ISPs. It won't happen --- getting new PA will be
> almost free, and getting PI address space is a nominal charge given that
> you have $$$$ for that bribe.
> 
Getting is cheap. Deploying is another matter.  It is the desire to avoid
the cost of deployment that will drive this, not the cost of the address
space. Leo summed up one likely scenario rather well.

>    michael> Everybody has a link local adddress. Things that are only
> 
> Alas, link-local addresses are not as easy to use as you might think.
> They are specific to a link, and IPv6 implementations insist that you
> tell the kernel which link they are specific to, so applications
> actually need to grow additional mechanisms to set that.
> Some OSes have included textual representations in their pton routines
> to set this properly, not it's not universal in the APIs.
> 
Riddle me this... When you have a Link-Local, ULA-C, and a GUA
address, how does one decide when to originate a connection from
the GUA and when to use the ULA-C address?

Until you have an answer for that that does not require updating
applications, the multiple different-capability addresses per
interface is a nice theory that is fraught with operational peril.

> It's a rare organization (or even residential home with a wifi router)
> that has only one physical link --- many do extensive amounts of L2
> bridging (and then filtering of broadcasts) to deal with scarcity of
> IPv4, and lack of mobility. 
> 
While that's true, it's hardly relevant to the topic at hand.

> In IPv6 it really makes no sense to do this --- make them seperate L3
> subnets and use properly scoped multicast to do service locating.
> 
Sure, but, that still doesn't change the overall issue of ULA-C.

Owen