One document matched: draft-ymbk-opcode-discover-01.txt
Differences from draft-ymbk-opcode-discover-00.txt
Individual Submission Bill Manning
draft-ymbk-opcode-discover-01.txt ISI
03 Jun 2001
The DISCOVER opcode
This document is an Internet-Draft and is NOT offered in accordance with
Section 10 of RFC2026, and the author does not provide the IETF with any
rights other than to publish as an Internet-Draft. This document is a
submission to the domain name system extentions (DNSEXT) working group
of the Internet Engineering Task Force (IETF). Comments may be submitted
to the working group mailing list at "namedroppers@ops.ietf.org" or the
author. Distribution of this memo is unlimited.
Internet-Drafts are working documents of the Internet Engineering Task
Force (IETF), its areas, and its working groups. Note that other groups
may also distribute working documents as Internet-Drafts.
Internet-Drafts are draft documents valid for a maximum of six months and
may be updated, replaced, or obsoleted by other documents at any time. It
is inappropriate to use Internet-Drafts as reference material or to cite
them other than as "work in progress."
The list of current Internet-Drafts can be accessed at
http://www.ietf.org/ietf/1id-abstracts.txt
The list of Internet-Draft Shadow Directories can be accessed at
http://www.ietf.org/shadow.html.
This work was funded under DARPA grant: F30602-99-1-0523
0. Abstract
The QUERY opcode in the DNS is designed for unicast. With the
development of multicast capabilities in the DNS, it is desireable
to have a more robust opcode for server interactions.
1. DISCOVER works like QUERY except:
1. it can be sent to a broadcast or multicast destination (QUERY
isn't defined for non-unicast, and arguably shouldn't be.)
2. the Question section, if present, has <QNAME=zonename,QTYPE=SOA>
tuples. Future work could augment this structure as follows:
<QNAME=service,QTYPE=SRV>
3. if QDCOUNT==0 then only servers willing to do recursion should
answer. Other servers must silently discard the DISCOVER request.
4. if QDCOUNT!=0 then only servers who are authoritative for the
zones named by some QNAME should answer.
5. responses may echo the request's Question section or leave it blank.
6. responses have "normal" Answer, Authority, and Additional sections.
e.g. the response is the same as that to a QUERY.
Usage for gethostby{name,addr}-style requestors:
Compute the zone name of the enclosing in-addr.arpa or ip6.int domain.
DISCOVER whether anyone in-scope is authoritative for this zone.
If so, query these authoritative servers for local
in-addr/ip6 names.
If not, DISCOVER whether there are recursive servers available.
If so, query these recursive servers for local
in-addr/ip6 names.
So, a node will issue a multicast request with the DISCOVER opcode at
some particular multicast scope. Then determine, from the replies,
whether there are any DNS servers which are authoritative (or support
recursion) for the zone.
Once one learns a host's FQDN by the above means, repeat the process
for discovering the closest enclosing authoritative server of such
local name.
Cache all NS and A data learned in this process, respecting TTL's.
Usage for SRV requestors:
Do the gethostbyaddr() and gethostbyname() on one's own LAN-local
address, using the above process.
Assume that the closest enclosing zone for which an authority server
answers an in-scope DISCOVER packet is "this host's parent domain".
Compute the SRV name as _service._transport.*.parentdomain.
This is a change to the definition as defined in RFC 1034.
A wildcard label ("*") in the QNAME used in a DNS message with
opcode DISCOVER SHOULD be evaluated with special rules. The
wildcard matches any label for which the DNS server data is
authoritative. For example 'x.*.example.com.' would match
'x.y.example.com.' and 'x.yy.example.com.' provided that the
server was authoritative for 'example.com.' In this particular
case, we suggest the follwing considerations be made:
getservbyname() can be satisfied by issuing a request with
this computed SRV name. The servent structure can be
populated by values returned from a request as follows:
s_name The name of the service, "_service" without the
preceding underscore.
s_aliases The names returned in the SRV RRs in replies
to the query.
s_port The port number in the SRV RRs replies to the
query. If these port numbers disagree - one
of the port numbers is chosen, and only those
names which correspond are returned.
s_proto The transport protocol from named by the
"_transport" label, without the preceding
underscore.
Send SRV query for this name to discovered local authority servers.
Usage for disconnected networks with no authority servers:
Hosts should run a "stub server" which acts as though its FQDN is a
zone name. Computed SOA gives the host's FQDN as MNAME, "." as the
ANAME, seconds-since-1Jan2000 as the SERIAL, low constants for EXPIRE
and the other timers. Computed NS gives the host's FQDN. Computed
glue gives the host's LAN-local address. Or Hosts may run a
"DNS stub server" which acts as though its FQDN is a zone name. The
rules governing the behavior of this stub server are given elsewhere
[1] [2].
Such stub servers should answer DISCOVER packets for its zone, and
will be found by the iterative "discover closest enclosing authority
server" by DISCOVER clients, either in the gethostbyname() or SRV
cases described above. Note that stub servers only answer with
zone names which match QNAME's, not with zone names which are owned
by QNAME's.
The only deviation from the DNS model is that a host (like, say, a printer
offering LPD services) has a DNS server which answers authoritatively for
something which hasn't been delegated to it. However, the only way that
such DNS servers can be discovered is with a new opcode, DISCOVER, which
is explicitly defined to discover undelegated zones for tightly scoped
purposes. Therefore this isn't officially a violation of DNS's coherency
principles.
The capitalized keywords "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in RFC 2119
3. IANA Considerations
As a new opcode, the IANA will need to assign a numeric value
for the memnonic.
4. Security Considerations
No new security considerations are known to be introduced. Using
multicast for service discovery has the potential for denial of
service.
5. Attribution:
This material was generated in discussions on the mdns mailing list
hosted by Zocalo in March 2000. Paul Vixie crystalized the concepts...
Stuart Cheshire, Bill Woodcock, Erik Guttman and were active contributors.
6. Author's Address
Bill Manning
PO 12317
Marina del Rey, CA. 90295
+1.310.322.8102
bmanning@karoshi.com
7. References
[1] draft-ietf-dnsext-mdns-00.txt
[2] draft-manning-dnsext-mdns-00.txt
| PAFTECH AB 2003-2026 | 2026-04-24 04:08:23 |