[Search] [pdf|bibtex] [Tracker] [WG] [Email] [Diff1] [Diff2] [Nits]

Versions: 00 01 02 03 04 05 06 07 08 rfc5507                            
Network Working Group                                                IAB
Internet-Draft                                         P. Faltstrom, Ed.
Intended status: Standards Track                         R. Austein, Ed.
Expires: August 21, 2008                                    P. Koch, Ed.
                                                       February 18, 2008


                   Design Choices When Expanding DNS
                        draft-iab-dns-choices-05

Status of this Memo

   By submitting this Internet-Draft, each author represents that any
   applicable patent or other IPR claims of which he or she is aware
   have been or will be disclosed, and any of which he or she becomes
   aware will be disclosed, in accordance with Section 6 of BCP 79.

   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 Internet-Draft will expire on August 21, 2008.

Copyright Notice

   Copyright (C) The IETF Trust (2008).

Abstract

   This note discusses how to extend the DNS with new data for a new
   application.  DNS extension discussions too often focus on reuse of
   the TXT Resource Record Type.  This document lists different
   mechanisms to extend the DNS, and concludes that the use of a new DNS
   Resource Record Type is the best solution.





IAB, et al.              Expires August 21, 2008                [Page 1]


Internet-Draft      Design Choices When Expanding DNS      February 2008


Table of Contents

   1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  3
   2.  Background . . . . . . . . . . . . . . . . . . . . . . . . . .  4
   3.  Extension mechanisms . . . . . . . . . . . . . . . . . . . . .  5
     3.1.  Place selectors inside the RDATA of existing Resource
           Record Types . . . . . . . . . . . . . . . . . . . . . . .  5
     3.2.  Add a prefix to the owner name . . . . . . . . . . . . . .  5
     3.3.  Add a suffix to the owner name . . . . . . . . . . . . . .  6
     3.4.  Add a new Class  . . . . . . . . . . . . . . . . . . . . .  7
     3.5.  Add a new Resource Record Type . . . . . . . . . . . . . .  7
   4.  Zone boundaries are invisible to applications  . . . . . . . .  8
   5.  Why adding a new Resource Record Type is the preferred
       solution . . . . . . . . . . . . . . . . . . . . . . . . . . .  9
   6.  Conclusion and Recommendation  . . . . . . . . . . . . . . . . 12
   7.  Creating A New Resource Record Type  . . . . . . . . . . . . . 13
   8.  IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 13
   9.  Security Considerations  . . . . . . . . . . . . . . . . . . . 13
   10. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 13
   11. References . . . . . . . . . . . . . . . . . . . . . . . . . . 14
     11.1. Normative References . . . . . . . . . . . . . . . . . . . 14
     11.2. Informative References . . . . . . . . . . . . . . . . . . 14
   Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 15
   Intellectual Property and Copyright Statements . . . . . . . . . . 16



























IAB, et al.              Expires August 21, 2008                [Page 2]


Internet-Draft      Design Choices When Expanding DNS      February 2008


1.  Introduction

   The DNS stores multiple categories of data.  The two most commonly
   used categories are infrastructure data for the DNS system itself (NS
   and SOA Resource Records) and data which have to do with mappings
   between domain names and IP addresses (A, AAAA and PTR Resource
   Records).  There are other categories as well, some of which are tied
   to specific applications like email (MX Resource Records), while
   others are generic Resource Record Types used to convey information
   for multiple protocols (SRV and NAPTR Resource Records).

   When storing data in the DNS for a new application, the data are
   usually tied to a "normal" domain name, so that the application can
   query for the data it wants, while minimizing the impact on existing
   applications.

   Historically, extending DNS to store application data tied to a
   domain name has been done in different ways at different times.  MX
   Resource Records were created as a new Resource Record Type
   specifically designed to support electronic mail.  SRV records are a
   generic type which use a prefixing scheme in combination with a base
   domain name.  Records associated with ENUM use a suffixing scheme.
   NAPTR records add selection data inside the RDATA.  It is clear that
   the methods used to add new data types to the DNS have been
   inconsistent, and the purpose of this document is to attempt to
   clarify the implications of each of these methods, both for the
   applications that use them and for the rest of the DNS.

   This document talks extensively about use of DNS wildcards.  Many
   people might think use of wildcards is not something that happens
   today.  In reality though, wildcards are in use, especially for
   certain application-specific data such as MX Resource Records.
   Because of this, the choice has to be made with existence of
   wildcards in mind.

   Another overall issue that must be taken into account is what the new
   data in the DNS are to describe.  In some cases they might be
   completely new data.  In other cases they might be metadata tied to
   data that already exist in the DNS.  An example of new data is key
   information for SSH and data used for fighting spam (metadata tied to
   MX Resource Records).  If the new data are tied to data that already
   exist in the DNS, an analysis should be made as to whether having
   (for example) address records and SSH key information in different
   DNS zones is a problem, and if it is, whether the specification must
   require all of the related data to be in the same zone.

   This document does not talk about what one should store in the DNS.
   It also doesn't discuss whether DNS should be used for service



IAB, et al.              Expires August 21, 2008                [Page 3]


Internet-Draft      Design Choices When Expanding DNS      February 2008


   discovery, or whether DNS should be used for storage of data specific
   for the service.  In general, DNS is a protocol that, apart from
   holding metadata that makes the DNS itself function (NS, SOA, DNSSEC
   Resource Record Types, etc), only holds references to service
   locations (SRV, NAPTR, A, AAAA Resource Record Types), but there are
   exceptions (such as MX Resource Records).


2.  Background

   See [RFC2929] for a brief summary of DNS query structure.  Readers
   interested in the full story should start with the base DNS
   specification in [RFC1035], and continue with the various documents
   that update, clarify, and extend the base specification.

   When composing a DNS query, the parameters used by the protocol are a
   triple: a DNS name, a DNS class, and a DNS record Type.  Every
   Resource Record matching a particular name, type and class is said to
   belong to the same "RRSet", and the whole RRSet is always returned to
   the client that queries for it.  Splitting an RRSet is a protocol
   violation, because it results in coherency problems with the DNS
   caching mechanism.

   Some discussions around extensions to the DNS include arguments
   around MTU size.  Note that most discussions about DNS and MTU size
   are about the size of the whole DNS packet, not about the size of a
   single RRSet.

   Almost all DNS query traffic is carried over UDP, where a DNS message
   must fit within a single UDP packet.  DNS response messages are
   almost always larger than DNS query messages, so message size issues
   are almost always about responses, not queries.  The base DNS
   specification limits DNS messages over UDP to 512 octets; EDNS0
   [RFC2671] specifies a mechanism by which a client can signal its
   willingness to receive larger responses, but deployment of EDNS0 is
   not universal, in part because of firewalls that block fragmented UDP
   packets or EDNS0.  If a response message won't fit in a single
   packet, the name server returns a truncated response, at which point
   the client may retry using TCP.  DNS queries over TCP are not subject
   to this length limitation, but TCP imposes significantly higher per-
   query overhead on name servers than UDP.  It is also the case that
   the policies in deployed firewalls far too often is such that it
   blocks DNS over TCP, so using TCP might not in reality be an option.
   There are also risks (although possibly small) that a change of
   routing while a TCP flow is open create problems when the DNS servers
   are deployed in an anycast environment.





IAB, et al.              Expires August 21, 2008                [Page 4]


Internet-Draft      Design Choices When Expanding DNS      February 2008


3.  Extension mechanisms

   The DNS protocol is intended to be extensible to support new kinds of
   data.  This section examines the various ways in which this sort of
   extension can be accomplished.

3.1.  Place selectors inside the RDATA of existing Resource Record Types

   For a given query name, one might choose to have a single RRSet (all
   Resource Records sharing the same name, type, and class) shared by
   multiple applications, and have the different applications use
   selectors within the Resource Record data (RDATA) to determine which
   records are intended for which applications.  This sort of selector
   mechanism is usually referred to "subtyping", because it is in effect
   creating an additional type subsystem within a single DNS Resource
   Record Type.

   Examples of subtyping include NAPTR Resource Records (see [RFC3761])
   and the original DNSSEC KEY Resource Record Type ([RFC2535]) (before
   it was updated by [RFC3445]).

   All DNS subtyping schemes share a common weakness: With subtyping
   schemes it is impossible for a client to query for just the data it
   wants.  Instead, the client must fetch the entire RRSet, then select
   the Resource Records in which it is interested.  Furthermore, since
   DNSSEC signatures operate on complete RRSets, the entire RRSet must
   be re-signed if any Resource Record in it changes.  As a result, each
   application that uses a subtyped Resource Record incurs higher
   overhead than any of the applications would have incurred had they
   not been using a subtyping scheme.  The fact the RRSet is always
   passed around as an indivisible unit increases the risk the RRSet
   will not fit in a UDP packet, which in turn increases the risk that
   the client will have to retry the query with TCP, which substantially
   increases the load on the name server.  More precisely: having one
   query fail over to TCP is not a big deal, but since the typical ratio
   of clients to servers in today's deployed DNS is very high, having a
   substantial number of DNS messages fail over to TCP may cause the
   queried name servers to be overloaded by TCP overhead.

   Because of the size limitations, using a subtyping scheme to list a
   large number of services for a single domain name risks triggering
   truncation and fallback to TCP, which may in turn force the zone
   administrator to announce only a subset of available services.

3.2.  Add a prefix to the owner name

   By adding an application-specific prefix to a domain name, we get a
   different name/class/type triple, and therefore a different RRSet.



IAB, et al.              Expires August 21, 2008                [Page 5]


Internet-Draft      Design Choices When Expanding DNS      February 2008


   One problem with adding prefixes has to do with wildcards, especially
   if one has records like

   *.example.com. IN MX 1 mail.example.com.

   and one wants records tied to those names.  Suppose one creates the
   prefix "_mail".  One would then have to say something like

   _mail.*.example.com. IN X-FOO A B C D

   but DNS wildcards only work with the "*" as the leftmost token in the
   domain name (see also [RFC4592]).

   Even when a specific prefix is chosen, the data will still have to be
   stored in some Resource Record Type.  This Resource Record Type can
   either be a record Type that has an appropriate format to store the
   data or a new Resource Record Type.  This implies that some other
   selection mechanism has to be applied as well, such as ability to
   distinguish between the records in an RRSet given they have the same
   Resource Record Type.  Because of this, one needs to both register a
   unique prefix and define what Resource Record Type is to be used for
   this specific service.

   If the record has some relationship with another record in the zone,
   the fact that the two records can be in different zones might have
   implications on the trust the application has in the records.  For
   example:

   example.com.      IN MX    10 mail.example.com.
   _foo.example.com. IN X-BAR "metadata for the mail service"

   In this example, the two records might be in two different zones, and
   because of this might be signed by two different organizations when
   using DNSSEC.

3.3.  Add a suffix to the owner name

   Adding a suffix to a domain name changes the name/class/type triple,
   and therefore the RRSet.  In this case, since the query name can be
   set to exactly the data one wants the size of the RRSet is minimized.
   The problem with adding a suffix is that it creates a parallel tree
   within the IN class.  Further, there is no technical mechanism to
   ensure that the delegation for "example.com" and "example.com._bar"
   are made to the same organization.  Furthermore, data associated with
   a single entity will now be stored in two different zones, such as
   "example.com" and "example.com._bar", which, depending on who
   controls "_bar", can create new synchronization and update
   authorization issues.



IAB, et al.              Expires August 21, 2008                [Page 6]


Internet-Draft      Design Choices When Expanding DNS      February 2008


   One way of solving the administrative issues is by using the DNAME
   Resource Record Type specified in [RFC2672].

   Even when using a different name, the data will still have to be
   stored in some Resource Record Type.  This Resource Record Type can
   either be a "kitchen-sink record" or a new Resource Record Type.
   This implies that some other mechanism has to be applied as well,
   with implications detailed in other parts of this note.

   In [RFC2163] an infix token is inserted directly below the TLD, but
   the result is equivalent to adding a suffix to the owner name
   (instead of creating a TLD one is creating a second level domain).

3.4.  Add a new Class

   DNS zones are class-specific in the sense that all the records in
   that zone share the same class as the zone's SOA record and the
   existence of a zone in one class does not guarantee the existence of
   the zone in any other class.  In practice, only the IN class has ever
   seen widespread deployment, and the administrative overhead of
   deploying an additional class would almost certainly be prohibitive.

   Nevertheless, one could in theory use the DNS class mechanism to
   distinguish between different kinds of data.  However, since the DNS
   delegation tree (represented by NS Resource Records) is itself tied
   to a specific class, attempting to resolve a query by crossing a
   class boundary may produce unexpected results because there is no
   guarantee that the name servers for the zone in the new class will be
   the same as the name servers in the IN class.  The MIT Hesiod system
   used a scheme like this for storing data in the HS class, but only on
   a very small scale (within a single institution), and with an
   administrative fiat requiring that the delegation trees for the IN
   and HS trees be identical.

   Even when using a different class, the data will still have to be
   stored in some Resource Record Type or another.  This Resource Record
   Type can either be a "kitchen-sink record" or a new Resource Record
   Type.  This implies that some other mechanism has to be applied as
   well, with implications detailed in other parts of this note.

3.5.  Add a new Resource Record Type

   When adding a new Resource Record Type to the system, entities in
   four different roles have to be able to handle the new Type:

   1.  There must be a way to insert the new Resource Records into the
       zone of the Primary Master name server.  For some server
       implementations, the user interface only accepts record Types



IAB, et al.              Expires August 21, 2008                [Page 7]


Internet-Draft      Design Choices When Expanding DNS      February 2008


       which it understands (perhaps so that the implementation can
       attempt to validate the data).  Other implementations allow the
       zone administrator to enter an integer for the Resource Record
       Type code and the RDATA in Base64 or hexadecimal encoding (or
       even as raw data).  [RFC3597] specifies a standard generic
       encoding for this purpose.
   2.  A slave authoritative name server must be able to do a zone
       transfer, receive the data from some other authoritative name
       server, and serve data from the zone even though the zone
       includes records of unknown Types.  Historically, some
       implementations have had problems parsing stored copies of the
       zone file after restarting, but those problems have not been seen
       for a few years.
   3.  A caching resolver (most commonly a recursive name server) will
       cache the records which are responses to queries.  As mentioned
       in [RFC3597],there are various pitfalls where a recursive name
       server might end up having problems.
   4.  The application must be able to get the RRSet with a new Resource
       Record Type.  The application itself may understand the RDATA,
       but the resolver library might not.  Support for a generic
       interface for retrieving arbitrary DNS Resource Record Types has
       been a requirement since 1989 (see [RFC1123] Section 6.1.4.2).
       Some stub resolver library implementations neglect to provide
       this functionality and cannot handle unknown Resource Record
       Types, but implementation of a new stub resolver library is not
       particularly difficult, and open source libraries that already
       provide this functionality are available.


4.  Zone boundaries are invisible to applications

   Regardless of the possible choices above we have seen a number of
   cases where the application made assumptions about the structure of
   the namespace and the location where specific information resides.
   We take a small sidestep to argue against such approaches.

   The DNS namespace is a hierarchy, technically speaking.  However,
   this only refers to the way names are built from multiple labels.
   DNS hierarchy neither follows nor implies administrative hierarchy.
   That said, it cannot be assumed that data attached to a node in the
   DNS tree is valid for the whole subtree.  Technically, there are zone
   boundaries partitioning the namespace and administrative boundaries
   (or policy boundaries) may even exist elsewhere.

   The false assumption has lead to an approach called "tree climbing",
   where a query that does not receive a positive response (either the
   requested RRSet was missing or the name did not exist) is retried by
   repeatedly stripping off the leftmost label (climbing towards the



IAB, et al.              Expires August 21, 2008                [Page 8]


Internet-Draft      Design Choices When Expanding DNS      February 2008


   root) until the root domain is reached.  Sometimes these proposals
   try to avoid the query for the root or the TLD level, but still this
   approach has severe drawbacks:

   o  Technically, the DNS was built as a query - response tool without
      any search capability [RFC3467].  Adding the search mechanism
      imposes additional burden on the technical infrastructure, in the
      worst case on TLD and root name servers.
   o  For reasons similar to those outlined in RFC 1535, querying for
      information in a domain outside the control of the intended entity
      may lead to incorrect results and may also put security at risk.
      Finding the exact policy boundary is impossible without an
      explicit marker which does not exist at present.  At best,
      software can detect zone boundaries (e.g., by looking for SOA
      Resource Records), but some TLD registries register names starting
      at the second level (e.g., CO.UK), and there are various other
      "registry" types at second, third or other level domains that
      cannot be identified as such without policy knowledge external to
      the DNS.

   To restate, the zone boundary is purely a boundary that exists in the
   DNS for administrative purposes, and applications should be careful
   not to draw unwarranted conclusions from zone boundaries.  A
   different way of stating this is that the DNS does not support
   inheritance, e.g. a wildcard MX RRSet for a TLD will not be valid for
   any subdomain of that particular TLD.


5.  Why adding a new Resource Record Type is the preferred solution

   By now, the astute reader might be be wondering what conclusions to
   draw from all the issues presented so far.  We will now attempt to
   clear up the reader's confusion by following the thought processes of
   a typical application designer who wishes to store data in the DNS,
   showing how such a designer almost inevitably hits upon the idea of
   just using TXT Resource Record, why this is a bad thing, and why a
   new Resource Record Type should be allocated instead.

   The overall problem with most solutions has to do with two main
   issues:
   o  No semantics to prevent collision with other use
   o  Space considerations in the DNS message

   A typical application designer is not interested in the DNS for its
   own sake, but rather as a distributed database in which application
   data can be stored.  As a result, the designer of a new application
   is usually looking for the easiest way to add whatever new data the
   application needs to the DNS in a way that naturally associates the



IAB, et al.              Expires August 21, 2008                [Page 9]


Internet-Draft      Design Choices When Expanding DNS      February 2008


   data with a DNS name.

   As explained in Section 3.4, using the DNS class system as an
   extension mechanism is not really an option, and in fact most users
   of the system don't even realize that the mechanism exists.  As a
   practical matter, therefore any extension is likely to be within the
   IN class.

   Adding a new Resource Record Type is the technically correct answer
   from the DNS protocol standpoint (more on this below), but doing so
   requires some DNS expertise, due to the issues listed in Section 3.5.
   As a result, this option is usually not considered.  Note that
   according to [RFC2929], some Types require IETF Consensus, while
   others only require a specification.

   The application designer is thus left with the prospect of reusing
   some existing DNS Type within the IN class, but when the designer
   looks at the existing Types, almost all of them have well-defined
   semantics, none of which quite match the needs of the new
   application.  This has not completely prevented proposals from
   reusing existing Resource Record Types in ways incompatible with
   their defined semantics, but it does tend to steer application
   designers away from this approach.

   For example, Resource Record Type 40 was registered for the SINK
   record Type.  This Resource Record Type was discussed in the DNSIND
   working group of the IETF, and it was decided at the 46th IETF to not
   move the I-D forward to become an RFC because of the risk of
   encouraging application designers to use the SINK Resource Record
   Type instead of registering a new Resource Record Type, which would
   result in infeasibly large SINK RRsets.

   Eliminating all of the above leaves the TXT Resource Record Type in
   the IN class.  The TXT RDATA format is free form text, and there are
   no existing semantics to get in the way.  Furthermore, the TXT
   Resource Record can obviously just be used as a bucket in which to
   carry around data to be used by some higher level parser, perhaps in
   some human readable programming or markup language.  Thus, for many
   applications, TXT Resource Records are the "obvious" choice.
   Unfortunately, this conclusion, while understandable, is also wrong,
   for several reasons.

   The first reason why TXT Resource Records are not well suited to such
   use is precisely the lack of defined semantics that make them so
   attractive.  Arguably, the TXT Resource Record is misnamed, and
   should have been called the Local Container record, because the lack
   of defined semantics means that a TXT Resource Record means precisely
   what the data producer says it means.  This is fine, so long as TXT



IAB, et al.              Expires August 21, 2008               [Page 10]


Internet-Draft      Design Choices When Expanding DNS      February 2008


   Resource Records are being used by human beings or by private
   agreement between data producer and data consumer.  However, it
   becomes a problem once one starts using them for standardized
   protocols in which there is no prior relationship between data
   producer and data consumer.  The reason for this is that there is
   nothing to prevent collisions with some other incompatible use of TXT
   Resource Records.  This is even worse than the general subtyping
   problem described in Section 3.1, because TXT Resource Records don't
   even have a standardized selector field in which to store the
   subtype.  [RFC1464] tried, but it was not a success.  At best a
   definition of a subtype is reduced to hoping that whatever scheme one
   has come up with will not accidently conflict with somebody else's
   subtyping scheme, and that it will not be possible to mis-parse one
   application's use of TXT Resource Records as data intended for a
   different application.  Any attempt to impose a standardized format
   within the TXT Resource Record format would be at least fifteen years
   too late even if it were put into effect immediately; at best, one
   can restrict the syntax that a particular application uses within a
   TXT Resource Record and accept the risk that unrelated TXT Resource
   Record uses will collide with it.

   Using one of the naming modifications discussed in Section 3.2 and
   Section 3.3 would address the subtyping problem, but each of these
   approaches brings in new problems of its own.  The prefix approach
   (that for example SRV Resource Records use) does not work well with
   wildcards, which is a particular problem for mail-related
   applications, since MX Resource Records are probably the most common
   use of DNS wildcards.  The suffix approach doesn't have wildcard
   issues, but, as noted previously, it does have synchronization and
   update authorization issues, since it works by creating a second
   subtree in a different part of the global DNS name space.

   The next reason why TXT Resource Records are not well suited to
   protocol use has to do with the limited data space available in a DNS
   message.  As alluded to briefly in Section 3.1, typical DNS query
   traffic patterns involve a very large number of DNS clients sending
   queries to a relatively small number of DNS servers.  Normal path MTU
   discovery schemes do little good here because, from the server's
   perspective, there isn't enough repeat traffic from any one client
   for it to be worth retaining state.  UDP-based DNS is an idempotent
   query, whereas TCP-based DNS requires the server to keep state (in
   the form of TCP connection state, usually in the server's kernel) and
   roughly triples the traffic load.  Thus, there's a strong incentive
   to keep DNS messages short enough to fit in a UDP datagram,
   preferably a UDP datagram short enough not to require IP
   fragmentation.

   Subtyping schemes are therefore again problematic because they



IAB, et al.              Expires August 21, 2008               [Page 11]


Internet-Draft      Design Choices When Expanding DNS      February 2008


   produce larger Resource RRSets than necessary, but verbose text
   encodings of data are also wasteful, since the data they hold can
   usually be represented more compactly in a Resource Record designed
   specifically to support the application's particular data needs.  If
   the data that need to be carried are so large that there is no way to
   make them fit comfortably into the DNS regardless of encoding, it is
   probably better to move the data somewhere else, and just use the DNS
   as a pointer to the data, as with NAPTR.


6.  Conclusion and Recommendation

   Given the problems detailed in Section 5, it is worth reexamining the
   oft-jumped-to conclusion that specifying a new Resource Record Type
   is hard.  Historically, this was indeed the case, but recent surveys
   suggest that support for unknown Resource Record Types [RFC3597] is
   now widespread, and that lack of support for unknown Types is mostly
   an issue for relatively old software that would probably need to be
   upgraded in any case as part of supporting a new application.  One
   should also remember that deployed DNS software today should support
   DNSSEC, and software recent enough to do so will likely support both
   unknown Resource Record Types [RFC3597] and EDNS0 [RFC2671].

   Of all the issues detailed in Section 3.5, provisioning the data is
   in some respects the most difficult.  The problem here is less
   difficult for the authoritative name servers themselves than the
   front-end systems used to enter (and perhaps validate) the data.
   Hand editing does not work well for maintenance of large zones, so
   some sort of tool is necessary, and the tool may not be tightly
   coupled to the name server implementation itself.  Note, however,
   that this provisioning problem exists to some degree with any new
   form of data to be stored in the DNS, regardless of data format,
   Resource Record type, or naming scheme.  Including the TXT Resource
   Record Type.  Adapting front-end systems to support a new Resource
   Record type may be a bit more difficult than reusing an existing
   type, but this appears to be a minor difference in degree rather than
   a difference in kind.

   Given the various issues described in this note, we believe that:
   o  there is no magic solution which allows a completely painless
      addition of new data to the DNS, but
   o  on the whole, the best solution is still to use the DNS Resource
      Record Type mechanism designed for precisely this purpose, and
   o  of all the alternate solutions, the "obvious" approach of using
      TXT Resource Records is almost certainly the worst.
   This especially for the two reasons outlined above (lack of semantics
   and its implications, and size leading to the need to use TCP).




IAB, et al.              Expires August 21, 2008               [Page 12]


Internet-Draft      Design Choices When Expanding DNS      February 2008


7.  Creating A New Resource Record Type

   The process for creating a new Resource Record Type is specified in
   [I-D.ietf-dnsext-2929bis].


8.  IANA Considerations

   This document does not require any IANA actions.


9.  Security Considerations

   DNS RRSets can be signed using DNSSEC.  DNSSEC is almost certainly
   necessary for any application mechanism that stores authorization
   data in the DNS.  DNSSEC signatures significantly increase the size
   of the messages transported, and because of this, the DNS message
   size issues discussed in Section 3.1 and Section 5 are more serious
   than they might at first appear.

   Adding new Resource Record Types (as discussed in Section 3.5) might
   conceivably trigger bugs and other bad behavior in software that is
   not compliant with [RFC3597], but most such software is old enough
   and insecure enough that it should be updated for other reasons in
   any case.  Basic API support for retrieving arbitrary Resource Record
   Types has been a requirement since 1989 (see [RFC1123]).

   Any new protocol that proposes to use the DNS to store data used to
   make authorization decisions would be well advised not only to use
   DNSSEC but also to encourage upgrades to DNS server software recent
   enough not to be riddled with well-known exploitable bugs.  Because
   of this, support for new Resource Record Types will not be as hard as
   people might think at first.


10.  Acknowledgements

   This document has been created during a number of years, with input
   from many people.  The question on how to expand and use the DNS is
   sensitive, and a document like this can not please everyone.  The
   goal is instead to describe the architecture and tradeoffs, and make
   some recommendations about best practices.

   People that have helped include: Dean Andersson, Loa Andersson, Mark
   Andrews, John Angelmo, Roy Badami, Dan Bernstein, Alex Bligh,
   Nathaniel Borenstein, Stephane Bortzmeyer, Brian Carpenter, Leslie
   Daigle, Elwyn Davies, Mark Delany, Richard Draves, Martin Duerst,
   Donald Eastlake, Robert Elz, Jim Fenton, Tony Finch, Jim Gilroy,



IAB, et al.              Expires August 21, 2008               [Page 13]


Internet-Draft      Design Choices When Expanding DNS      February 2008


   Olafur Gudmundsson, Eric Hall, Philip Hallam-Baker, Ted Hardie, Bob
   Hinden, Paul Hoffman, Geoff Houston, Christian Huitema, Johan Ihren,
   John Klensin, Olaf Kolkman, Ben Laurie, William Leibzon, John Levine,
   Edward Lewis, David MacQuigg, Allison Manking, Bill Manning, Danny
   McPherson, David Meyer, Pekka Nikander, Masataka Ohta, Douglas Otis,
   Michael Patton, Jonathan Rosenberg, Anders Rundgren, Miriam Sapiro,
   Pekka Savola, Chip Sharp, James Snell, Dave Thaler, Michael Thomas,
   Paul Vixie, Sam Weiler, Florian Weimer, Bert Wijnen, and Dan Wing.

   Members of the IAB when this document was made available were: Loa
   Andersson, Leslie Daigle, Elwyn Davies, Kevin Fall, Russ Housley,
   Olaf Kolkman, Barry Leiba, Kurtis Lindqvist, Danny McPherson, David
   Oran, Eric Rescorla, Dave Thaler, and Lixia Zhang.


11.  References

11.1.  Normative References

   [RFC1035]  Mockapetris, P., "Domain names - implementation and
              specification", RFC 1035, STD 13, November 1987.

   [RFC1464]  Rosenbaum, R., "Using the Domain Name System To Store
              Arbitrary String Attributes", RFC 1464, May 1993.

   [RFC2535]  Eastlake 3rd, D., "Domain Name System Security
              Extensions", RFC 2535, March 1999.

   [RFC2671]  Vixie, P., "Extension Mechanisms for DNS (EDNS0)",
              RFC 2671, August 1999.

   [RFC3597]  Gustafsson, A., "Handling of Unknown DNS Resource Record
              (RR) Types", RFC 3597, September 2003.

11.2.  Informative References

   [I-D.ietf-dnsext-2929bis]
              3rd, D., "Domain Name System (DNS) IANA Considerations",
              draft-ietf-dnsext-2929bis-06 (work in progress),
              August 2007.

   [RFC1123]  Braden, R., "Requirements for Internet Hosts - Application
              and Support", RFC 1123, STD 3, October 1989.

   [RFC2163]  Allocchio, C., "Using the Internet DNS to Distribute MIXER
              Conformant Global Address Mapping (MCGAM)", RFC 2163,
              January 1998.




IAB, et al.              Expires August 21, 2008               [Page 14]


Internet-Draft      Design Choices When Expanding DNS      February 2008


   [RFC2672]  Crawford, M., "Non-Terminal DNS Name Redirection",
              RFC 2672, August 1999.

   [RFC2929]  Eastlake 3rd, D., Brunner-Williams, E., and B. Manning,
              "Domain Name System (DNS) IANA Considerations", RFC 2929,
              BCP 42, September 2000.

   [RFC3445]  Massey, D. and S. Rose, "Limiting the Scope of the KEY
              Resource Record (RR)", RFC 3445, December 2002.

   [RFC3467]  Klensin, J., "Role of the Domain Name System (DNS)",
              RFC 3467, February 2003.

   [RFC3761]  Faltstrom, P. and M. Mealling, "The E.164 to Uniform
              Resource Identifiers (URI) Dynamic Delegation Discovery
              System (DDDS) Application (ENUM)", RFC 3761, April 2004.

   [RFC4592]  Lewis, E., "The Role of Wildcards in the Domain Name
              System", RFC 4592, July 2006.


Authors' Addresses

   Internet Architecture Board

   Email: iab@iab.org


   Patrik Faltstrom (editor)

   Email: paf@cisco.com


   Rob Austein (editor)

   Email: sra@isc.org


   Peter Koch (editor)

   Email: pk@denic.de










IAB, et al.              Expires August 21, 2008               [Page 15]


Internet-Draft      Design Choices When Expanding DNS      February 2008


Full Copyright Statement

   Copyright (C) The IETF Trust (2008).

   This document is subject to the rights, licenses and restrictions
   contained in BCP 78, and except as set forth therein, the authors
   retain all their rights.

   This document and the information contained herein are provided on an
   "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
   OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND
   THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS
   OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF
   THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED
   WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.


Intellectual Property

   The IETF takes no position regarding the validity or scope of any
   Intellectual Property Rights or other rights that might be claimed to
   pertain to the implementation or use of the technology described in
   this document or the extent to which any license under such rights
   might or might not be available; nor does it represent that it has
   made any independent effort to identify any such rights.  Information
   on the procedures with respect to rights in RFC documents can be
   found in BCP 78 and BCP 79.

   Copies of IPR disclosures made to the IETF Secretariat and any
   assurances of licenses to be made available, or the result of an
   attempt made to obtain a general license or permission for the use of
   such proprietary rights by implementers or users of this
   specification can be obtained from the IETF on-line IPR repository at
   http://www.ietf.org/ipr.

   The IETF invites any interested party to bring to its attention any
   copyrights, patents or patent applications, or other proprietary
   rights that may cover technology that may be required to implement
   this standard.  Please address the information to the IETF at
   ietf-ipr@ietf.org.


Acknowledgment

   Funding for the RFC Editor function is provided by the IETF
   Administrative Support Activity (IASA).





IAB, et al.              Expires August 21, 2008               [Page 16]