TR : Stepping back on the DNS discovery discussion

["BELOEIL Luc FTRD/DMI/CAE" <luc.beloeil@francetelecom.com>]

(sorry it seems I have encountered pb to send messages...)


Hi Alain, 

> -----Message d'origine-----
> De : Alain Durand [mailto:Alain.Durand@Sun.COM]
> Envoyé : mardi 15 juillet 2003 09:40

I agree, we should first have a clear consensus on requirements for "DNS discovery".

> A)  Rob, in his introduction, pointed at a number a functions a node
> could

I agree that point 1- is the most critical requirement for basic services. It's not clear for me how to evaluate criticity of other points.

> B) The second major element to study is the desired/required mode of
> operation.

As far as I know, 3 solutions have been documented:
- well-known site-local IPv6 addresses... IMHO, that solution depends on decision of ipv6 wg upon local addresses
- DHCPv6 (+ DHCP-lite)
- RA-based solution
Some guys had also spoken about SLPv6. But at last meeting, Bob pointed that a service discovery protocol should not be used for that (no one answered to that).

As DNS is a critical service between "applications world" and "IP layer world", we might be careful on consequences. From application point of view, we might ask in the future to extende that solution to other type of services (NTP, ...). In that way, RA-based solution might not be the good solution, because I do think that RA must not be extended to any sort of service discovery. Any idea about that?

> C) A third element in this discussion is the compatibility with the
> IPv4 model.

I agree that we need an _administrative model_ in IPv6. But I don't really understand that within that discussion. Do you have in mind O-bit et al? Anyway could you clarify?


> 	- Alain.
> 
> #-------------------------------------------------------------
> ---------
> # To unsubscribe, send a message to <dnsop-request@cafax.se>.
> 

#----------------------------------------------------------------------
# To unsubscribe, send a message to <dnsop-request@cafax.se>.