Internet Draft
Internet-Draft Ryan Moats
draft-ietf-svrloc-discovery-02.txt AT&T
Expires: December 1997 Martin Hamilton
Loughborough University
Paul J. Leach
Microsoft
June 1997
Finding Stuff
(How to discover services)
Filename: draft-ietf-svrloc-discovery-02.txt
Status of This Memo
This document is an Internet-Draft. 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.''
To learn the current status of any Internet-Draft, please check
the ``1id-abstracts.txt'' listing contained in the Internet-
Drafts Shadow Directories on ftp.is.co.za (Africa), nic.nordu.net
(Europe), munnari.oz.au (Pacific Rim), ds.internic.net (US East
Coast), or ftp.isi.edu (US West Coast).
Abstract
This document proposes a solution to the problem of finding
information about that services are being offered at a particular
Internet domain. Therefore, it is possible for clients, using this
approach, to locate services in a domain with only prior knowledge of
the domain name.
1. Rationale
Currently, there is no one single way of discovering the network
services and application protocols supported at a particular Internet
domain. The Domain Name System (DNS) provides some basic facilities
Expires 12/31/97 [Page 1]
INTERNET DRAFT Finding Stuff June 1997
for finding the hosts that offer particular services, such as DNS
servers themselves (NS records), mail exchangers (MX records [3]).
Recently general service records (SRV records [1]) have been proposed
for DNS, along with storing geographic information (LOC records [6]).
In addition, there are evolving methods for doing service location
via other methods [4] & [5].
This document sets forth a process to rationalize how a client could
use these various methods for service location. In this process, the
use of the Service Location Protocol [4] is highlighted to allow
clients to discover services by characteristics rather than by type
alone. With the use of SRVs alone, all services are assumed to be
identical except for weight. While this may be the case at some
locations, SLP will still be useful as it provides a dynamic
registration framework for services.
2. The process
This process is an aggregation of several different ideas. To
explicitly point out which ideas are being used, the steps of the
process have been grouped by section according to which technique is
being used.
For a client in domain "srcdom" that wants to locate service
"service" in domain "tgtdom", the following process should be
followed:
2.1 Support for "Sites" Alias Structure (see Appendix A)
1. If "srcdom" is of the form "@" and
"" is a suffix of "tgtdom", then do steps 2
through 9 with "target" set to ".sites.". If
this procedure fails or the condition is not met, set "target" to
"tgtdom".
2.2 Support for SVRLOC
2. If "srcdom" == "target" and the client supports it, use the
Service Location Protocol to determine if the service can be found
that way.
3. If the client supports Service Location Protocol, look for SRV
records [1] for a directory agent (i.e. da.udp."target" or
da.tcp."target"). This consists of:
3a. Do a lookup of QNAME=da.tcp.target, QCLASS=IN, QTYPE=SRV.
Expires 12/31/97 [Page 2]
INTERNET DRAFT Finding Stuff June 1997
3b. If the reply is NOERROR, ANCOUNT>0 and there is at least one
SRV RR that specifies the requested Service and Protocol in the
reply:
If there is precisely one SRV RR, and its Target is "." (the
root domain), go to step 3c.
For all SRV RR's build a list of (Priority, Weight, Target)
tuples and sort the list by priority (lowest number first).
Create a new empty list.
For each distinct priority level:
For each element at this priority level:
While there are still elements left at this priority level,
select an element randomly, with probability weight, and
move it to the tail of the new list.
For each element in the new list:
Query the DNS for A RR's for the target or use any RR's
found in the additional data section of the earlier SRV
query.
For each A RR found, try to connect to the directory agent
using the Service Location Protocol over TCP.
3c. Else do a lookup of QNAME=da.udp.target, QCLASS=IN, QTYPE=SRV.
3d. Process the reply as in 3b, except that if there is precisely
one SRV RR with a target of ".", go to step 4 and connections to
the directory agent use Service Location Protocol over UDP.
2.3 Support for SRV records
4. Look for SRV records for service.protocol.target, where
protocol is whichever protocol (TCP or UDP) is associated with
service. This consists of:
4a. Do a lookup for QNAME=service.protocol.target, QCLASS=IN,
QTYPE=SRV.
4b. If the reply is NOERROR, ANCOUNT>0 and there is at least one
SRV RR that specifies the requested Service and Protocol in the
reply:
If there is precisely one SRV RR, and its Target is "." (the
root domain), go to step 5.
Expires 12/31/97 [Page 3]
INTERNET DRAFT Finding Stuff June 1997
For all SRV RR's, build a list of (Priority, Weight, Target)
tuples and sort the list by priority (lowest number first).
Create a new empty list.
For each distinct priority level:
For each element at this priority level:
Query the DNS for LOC RR [6] for the Target (if not found in
the Additional Data section of the earlier SRV query).
Find the nearest target and all targets "close" to the
nearest target (target to target distance less than 2-3% of
client to nearest target distance).
Remove all other targets at this priority level.
While there are still elements left at this priority level,
select an element randomly, with probability weight, and
move it to the tail of the new list.
For each element in the new list:
Query the DNS for A RR's for the Target or use any RR's
found in the Additional Data section of the earlier SRV
query.
For each A RR found, if the protocol is TCP (connection-
oriented) try to connect to the (protocol, address,
service); if the protocol is UDP, send a service request.
5. If the service desired is SMTP, skip to RFC 974 (MX records)
else go to step 6.
2.3 Support for "Well Known" Aliases
6. If the service has a "well known" alias (see [2]) service',
look for A RRs for service'.target. This is done in the following
way:
6a. Do a lookup for QNAME=service'.target, QCLASS=IN, QTYPE=A.
If no A RR's returned, go to step 7.
For each A RR found, try to connect to the (protocol, address,
service). If successful, stop.
If all A RR's have been tried go to step 7.
Expires 12/31/97 [Page 4]
INTERNET DRAFT Finding Stuff June 1997
2.4 Support for Service Advertising via Service URLs
7. Look for "service:" URLs stored in TXT RRs for service.target:
7a. Do a lookup for QNAME=service.target, QCLASS=IN, QTYPE=TXT.
If no TXT RR's returned, go to step 8.
For each TXT RR found, try to connect to the (address, port)
specified in the service: URL. If successful, stop.
If all TXT RR's have been tried go to step 8.
8. Look for "service:" URLs stored in TXT RRs for target:
8a. Do a lookup for QNAME=target, QCLASS=IN, QTYPE=TXT.
If no TXT RR's returned, go to step 9.
For each TXT RR found, try to connect to the (address, port)
specified in the service: URL. If successful, stop.
If all TXT RR's have been tried go to step 9.
2.5 Fallback
9. Do a lookup for QNAME=target, QCLASS=IN, QTYPE=A
For each A RR found, try to connect to the (protocol, address,
service). If successful, stop.
3. Security Considerations
Because of the suggested mechanisms for service discovery, this
document inherits all the security considerations of using DNS RR's
and the Service Location Protocol. Implementors should consider both
[7] and the security section of [4] for appropriate methods.
4. Conclusion
By following the above process, a client may be reasonably certain of
determining whether a particular service is provided for a particular
domain name, given the domain name.
Expires 12/31/97 [Page 5]
INTERNET DRAFT Finding Stuff June 1997
5. Acknowledgments
This document is partially supported by the National Science
Foundation, Cooperative Agreement NCR-9218179, the UK Electronic
Libraries Programme (eLib) grant 12/39/01, and the European
Commission's Telematics for Research Programme grant RE 1004.
6. References
Request For Comments (RFC) and Internet Draft documents are available
from and numerous mirror sites.
[1] A. Gulbrandsen, P. Vixie, "A DNS RR for specifying
the location of services (DNS SRV)," RFC 2052,
October 1996.
[2] M. Hamilton, R. Wright, "Use of DNS Aliases for
Network Services," Internet Draft (work in pro-
gress), June 1996.
[3] S. C. Partridge, "Mail routing and the domain sys-
tem," RFC 974, January 1, 1986.
[4] J. Veizades, E. Guttman, C. Perkins, S. Kaplan,
"Service Location Protocol," Internet Draft (work
in progress), April 3, 1997.
[5] R. Moats, M. Hamilton, "Advertising Services,"
Internet Draft (work in progress), February 1997.
[6] C. Davis, P. Vixie, T. Goodwin, I. Dickinson, "A
Means for Expressing Location Information in the
Domain Name System," RFC 1876, January 15, 1996.
[7] D. Eastlake, C. Kaufman, "Domain Name System Secu-
rity Extensions," RFC 2065, January 3, 1997.
Expires 12/31/97 [Page 6]
INTERNET DRAFT Finding Stuff June 1997
7. Authors' addresses
Ryan Moats
AT&T
15621 Drexel Circle
Omaha, NE 68135-2358
USA
Phone: +1 402 894-9456
EMail: jayhawk@ds.internic.net
Martin Hamilton
Department of Computer Studies
Loughborough University of Technology
Leics. LE11 3TU, UK
Email: m.t.hamilton@lut.ac.uk
Paul J. Leach
Microsoft
1 Microsoft Way
Redmond, Washington, 98052, U.S.A.
Email: paulle@microsoft.com
A. Discovery with Large numbers of replicas
Imagine an organization (such as a bank) which has thousands of
branch offices. For reliability, each office may want to have a
replica of (at least) certain critical information in case of loss of
connectivity to the Internet. An example of such information might be
an address book for employees or an authentication database. It is
not reasonable to use the method of the previous section to locate a
replica for the information - there would be thousands of SRV entries
for the service being replicated. Instead, it would be desirable to
have a method that first looked to see if there is a replica for the
service at the branch office, and only proceed with the more general
discovery method if there weren't.
Let us generalize the notion of branch office to that of "site" - a
site is a collection of hosts that have good enough connectivity that
use of a service instance at the site is always to be preferred to
one at another, and that there is no connectivity reason to prefer
one replica within a site to another. A site has a site name that
incorporates a site specific component and the domain name of the
organization of the form
@
Then a client at site
@
looking for a service for domain , where is a suffix of , should use the name
.sites.
as the value of "target" in the procedure described herein. If
the procedure fails, then it should try with
as the value of "target" using the procedure presented above. Note
that within a site, this means the client either uses SVRLOC (if
supported) or straight SRV records; by definition of "site", there
would be no advantage to be gained from using LOC records.
For example, suppose a client at the site "dublin@univexports.com"
wanted to access the LDAP server for the Asian region of Universal
Exports, whose domain name was "fareast.univexports.com". It would
observe that its of "univexports.com" was a suffix
of "fareast.univexports.com", and first build the name
"dublin.sites.fareast.univexports.com" to use as the "target" in the
procedure above; this might cause it to then lookup
SRV records for "ldap.tcp.dublin.sites.fareast.univexports.com" or
A records for "ldap.dublin.sites.fareast.univexports.com" or even
TXT records for "dublin.sites.fareast.univexports.com" (looking
for "service:" URLs of the type wp-ldap).
Note that the list of records for a site does not have to point
at only hosts at that site - for example, the highest priority
records could be for hosts at the site, and then some lower priority
records for hosts at the best-connected alternate site for backup.
Using this method, a client looking for a service that has an replica
at its site will only have to fetch records for its site, not for
the whole organization. The records associated with the unadorned
organization name can then be used to aid clients from outside the
organization, who have no idea of the site structure of the
organization.
This Internet Draft expires December 31, 1997.
Expires 12/31/97 [Page 8]