Re: [lamps] DNS DNAME pain.

[Phillip Hallam-Baker <[e-mail]>]

On Thu, Nov 9, 2017 at 11:53 AM, John R Levine <[e-mail]> wrote:
> On Thu, 9 Nov 2017, Phillip Hallam-Baker wrote:
>>>
>>> DNAME only maps names below it, so the CAA is fine.
>>
>>
>> That is my understanding from the text, the example suggests
>> otherwise. I would like to check that the old behavior of not matching
>> the root is still valid.
>
>
> It hasn't changed.  For a long time the .CAT domain made a poor attempt to
> map ASCII and accented versions of names using DNAME, and I found out all
> about DNAME+other stuff.
>
>>> There's a BNAME proposal kicking around that is sort of a combined
>>> CNAME and DNAME, mapping everything at and below the name.  Or do you
>>> mean something else like a translucent DNAME that only maps if there's
>>> nothing at the actual name?
>>
>>
>> I am aware that there have been proposals circulating, I have not been
>> tracking them directly. They seem to always be about to happen in a
>> year or two.
>
>
> BNAME is going nowhere because it doesn't solve the problem it's supposed to
> solve, mapping variant names together.  If you want, say, a pair of
> traditional and simplified Chinese names to act the same, you have far more
> work provisioning web and mail servers than the DNS.

Like many other DNS proposals, the only way that can work is if the
services are getting their configuration data from the same source of
truth as the authoritative DNS server

That is something I have been building for my own use. When a Docker
container pops up, it should register itself in the DNS as one of the
handlers for that service. When it dies, it should be unregistered.