Re: [mdnsext] Hierarchical (host) domain names in mDNS?

["Albrecht, Harald" <[e-mail]>]

Hi,

thanks for the valuable feedback. I have no problem with explicitly appending ".local.", in fact, I would even like to make this explicit, so that "local" connections also need to use "controller.machine1.local.". This way, users will learn right from the start which names will trigger which resolution technologies.

My (negative) experience comes from experimenting with the mDNS tools in the Apple *nix package (if I remember correctly, the exact tests did a colleague of mine). The problem was that these command line tools reject any host domain name that doesn't fit into the "single-dns-label.localhost" template. So that is where my question and hesitation comes from. Of course, we may have done something wrong and the error is on our side. Anyway, I'm curious to learn more here in order to better understand the full picture.

So, from what you wrote I will ask my colleague to run the checks again not using the command line tools but using the mDNS resolver instead. It will be interesting what the results will be... looks as if there is a difference between the resolver and the command line tools.

With best regards,
Harald

Siemens AG
Industry Sector
Industry Automation Division
Industrial Automation Systems
I IA AS CTO DH 1
Gleiwitzer Str. 555
90475 Nürnberg, Deutschland
Tel: [phone]
Fax: [phone]
mailto:[e-mail]

Siemens Aktiengesellschaft: Vorsitzender des Aufsichtsrats: Gerhard Cromme; Vorstand: Peter Löscher, Vorsitzender; Roland Busch, Brigitte Ederer, Klaus Helmrich, Joe Kaeser, Barbara Kux, Hermann Requardt, Siegfried Russwurm, Peter Y. Solmssen, Michael Süß; Sitz der Gesellschaft: Berlin und München, Deutschland; Registergericht: Berlin Charlottenburg, HRB 12300, München, HRB 6684; WEEE-Reg.-Nr. DE 23691322



Von: [e-mail] [mailto:[e-mail]] Im Auftrag von Kerry Lynn
Gesendet: Mittwoch, 17. Juli 2013 16:13
An: Albrecht, Harald
Cc: [e-mail]
Betreff: Re: [mdnsext] Hierarchical (host) domain names in mDNS?

On Wed, Jul 17, 2013 at 3:13 AM, Albrecht, Harald <[e-mail]<mailto:[e-mail]>> wrote:
Thanks for the reply,

control over the software stack is right down to one of the problems: there is not a single software stack, there are multiple ones, so while we may have control over some of our own stacks, we won't have over the ones already out in the wild and which we want to interoperate with. Maybe a change would be possible, but my view is that this would first need consensus (where exactly? here in this group?) and in the "worst cast" an update to the RFC. Would this be a feasible approach at all?

Harald,

As I said previously, if your application provides a qualified domain name ending in
".local" or ".local." then what you propose should work fine with an mDNS-aware
resolver.  It is the TLD ".local" that triggers the behavior to multicast the request to
port 5353.  What an mDNS-aware resolver won't do is to append ".local" to a multi-
label host name if your application omits it.

IMO, this does not constitute "full control of the stack", nor is an update to RFC 6762
required.  This is not a difficult experiment to set up; I have tried it with OS X and
Ubuntu LTS.  You can ensure it works on Windows by installing Bonjour for Windows.

-K-

With best regards,
Harald Albrecht

Siemens AG
Industry Sector
Industry Automation Division
Industrial Automation Systems
I IA AS CTO DH 1
Gleiwitzer Str. 555
90475 Nürnberg, Deutschland
Tel: [phone]<tel:%2B49%20911%20895-3847>
Fax: [phone]<tel:%2B49%20911%20895-2105>
mailto:[e-mail]<mailto:[e-mail]>

Siemens Aktiengesellschaft: Vorsitzender des Aufsichtsrats: Gerhard Cromme; Vorstand: Peter Löscher, Vorsitzender; Roland Busch, Brigitte Ederer, Klaus Helmrich, Joe Kaeser, Barbara Kux, Hermann Requardt, Siegfried Russwurm, Peter Y. Solmssen, Michael Süß; Sitz der Gesellschaft: Berlin und München, Deutschland; Registergericht: Berlin Charlottenburg, HRB 12300, München, HRB 6684; WEEE-Reg.-Nr. DE 23691322


> -----Ursprüngliche Nachricht-----
> Von: [e-mail]<mailto:[e-mail]> [mailto:[e-mail]<mailto:[e-mail]>] Im
> Auftrag von Andrew Sullivan
> Gesendet: Dienstag, 16. Juli 2013 17:15
> An: [e-mail]<mailto:[e-mail]>
> Betreff: Re: [mdnsext] Hierarchical (host) domain names in mDNS?
>
> On Tue, Jul 16, 2013 at 10:49:11AM -0400, Kerry Lynn wrote:
>
> > the hen house.  One should still expect "controller.machine2.local."
> > to resolve properly,
>
> Yes, if you tried to resolve "controller.machine2.local.", then it _would_
> resolve correctly.  The difficulty, however, is that most resolving applications
> don't actually work truly fully-qualified that way (i.e. by appending the final
> dot).  So it will work fine if you have special name-handling code to make
> sure that everything is actually really fully qualified always in local.  I think it
> would be extremely wise to test how well that works in practice, but those
> who have complete control over their application stack and so on might be
> able to do it.
>
> Best,
>
> A
>
> --
> Andrew Sullivan
> [e-mail]<mailto:[e-mail]>
> _______________________________________________
> mdnsext mailing list
> [e-mail]<mailto:[e-mail]>
> https://www.ietf.org/mailman/listinfo/mdnsext
_______________________________________________
mdnsext mailing list
[e-mail]<mailto:[e-mail]>
https://www.ietf.org/mailman/listinfo/mdnsext