Document: draft-warnicke-network-dns-resolution-05.txt
From: Suzanne Woolf 
Date: 2 mars 2005 

This returns to my stack with a note from the AD:

"Note: 2005-02-23: I've reviewed this and do not believe it conflicts 
with
any IETF work. I think is fine to be published as an Independent
Submission."

When I reviewed this draft before, I suggested that DNSOP review it to
see if a usage note or other modification might be appropriate. I
didn't mean I thought there was a conflict as such, just that DNSOP
exists in significant measure to review submissions such as this, per
its charter item:

1. Define the processes by which Domain Name System (DNS) software
       may be efficiently and correctly administered, configured, and
       operated on Internet networks. This will include root zone
       name servers, gTLD name servers, name servers for other DNS
       zones, iterative DNS resolvers, and recursive DNS resolvers.
       As part of this effort, the group will produce documents
       explaining to the general Internet community what processes
       and mechanisms should be employed for the effective management
       and operation of DNS software.

The mechanism discussed in the Warnicke draft seems to me to a) have
potentially poor scaling or other operational characteristics and b)
to involve specific aspects of how the DNS may be "administered,
configured, and operated on Internet networks."

Sorry if I'm being dense, but I still like my original suggestion.  If
David as AD thinks the extra review step is unnecessary or
inappropriate, or the DNSOP chairs have rejected the work item, it's
not gonna happen....but I'll probably keep sitting here quietly
thinking it should. :)


Suzanne

Re:
> Document: draft-warnicke-network-dns-resolution-05.txt
> Title: "A Suggested Scheme for DNS Resolution of Networks and Gateways"
> Intended Status: listed as informational, see below
> Reviewer: Suzanne Woolf
> Review Date: 1 February 2004
> Submitted through RFC-editor
>
> The draft is listed in the agenda as "informational". As written, it
> may be intended as "BCP". But there are no usage notes revealing the
> intention.
>
> The draft discusses a mechanism for using specially-formed chains of
> PTR records to publish the names and IP addresses of the first-hop
> router for a given IP address, and (implicitly) the netmask. The
> mechanism is not too dissimilar to that used for subdelegating
> responsibility for PTR records in prefix delegations longer than /24
> (RFC 2317).
>
> At first reading, this strikes me as the sort of use of the DNS often
> proposed by people who think the DNS is a more sophisticated and
> robust database mechanism than it is. Such schemes tend to be harmless
> if people decide they're merely useless, but often they don't scale
> very well if people do like them. There are also may be implementation
> pitfalls that can generate lots of new queries on the infrastructure.
>
> I suggest this draft be reviewed by DNSOP, if only to add usage and
> implementation notes suggesting the limitations or hazards of the
> proposed approach. There are no protocol issues I can see so it's out
> of scope for DNSEXT.
>