RFC 8324






Independent Submission                                        J. Klensin
Request for Comments: 8324                                 February 2018
Category: Informational
ISSN: 2070-1721


    DNS Privacy, Authorization, Special Uses, Encoding, Characters,
          Matching, and Root Structure: Time for Another Look?

Abstract



   The basic design of the Domain Name System was completed almost 30
   years ago.  The last half of that period has been characterized by
   significant changes in requirements and expectations, some of which
   either require changes to how the DNS is used or can be accommodated
   only poorly or not at all.  This document asks the question of
   whether it is time to either redesign and replace the DNS to match
   contemporary requirements and expectations (rather than continuing to
   try to design and implement incremental patches that are not fully
   satisfactory) or draw some clear lines about functionality that is
   not really needed or that should be performed in some other way.

Status of This Memo



   This document is not an Internet Standards Track specification; it is
   published for informational purposes.

   This is a contribution to the RFC Series, independently of any other
   RFC stream.  The RFC Editor has chosen to publish this document at
   its discretion and makes no statement about its value for
   implementation or deployment.  Documents approved for publication by
   the RFC Editor are not candidates for any level of Internet Standard;
   see Section 2 of RFC 7841.

   Information about the current status of this document, any errata,
   and how to provide feedback on it may be obtained at
   https://www.rfc-editor.org/info/rfc8324.














Klensin                       Informational                     [Page 1]

RFC 8324                      DNS Revisions                February 2018


Copyright Notice



   Copyright (c) 2018 IETF Trust and the persons identified as the
   document authors.  All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents
   (https://trustee.ietf.org/license-info) in effect on the date of
   publication of this document.  Please review these documents
   carefully, as they describe your rights and restrictions with respect
   to this document.








































Klensin                       Informational                     [Page 2]

RFC 8324                      DNS Revisions                February 2018


Table of Contents



   1.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   4
   2.  Background and Hypothesis . . . . . . . . . . . . . . . . . .   5
   3.  Warts and Tensions with the Current DNS . . . . . . . . . . .   6
     3.1.  Multi-type Queries  . . . . . . . . . . . . . . . . . . .   6
     3.2.  Matching Part I: Case Sensitivity in Labels and Other
           Anomalies . . . . . . . . . . . . . . . . . . . . . . . .   7
     3.3.  Matching Part II: Non-ASCII ("Internationalized") Domain
           Name Labels . . . . . . . . . . . . . . . . . . . . . . .   7
     3.4.  Matching Part III: Label Synonyms, Equivalent Names, and
           Variants  . . . . . . . . . . . . . . . . . . . . . . . .   8
     3.5.  Query Privacy . . . . . . . . . . . . . . . . . . . . . .  10
     3.6.  Alternate Namespaces for Public Use in the DNS Framework:
           The CLASS Problem . . . . . . . . . . . . . . . . . . . .  10
     3.7.  Loose Synchronization . . . . . . . . . . . . . . . . . .  10
     3.8.  Private Namespaces and Special Names  . . . . . . . . . .  11
     3.9.  Alternate Query or Response Encodings . . . . . . . . . .  12
     3.10. Distribution and Management of Root Servers . . . . . . .  12
     3.11. Identifiers versus Brands and Other Convenience Names . .  13
     3.12. A Single Hierarchy with a Centrally Controlled Root . . .  14
     3.13. Newer Application Protocols, New Requirements, and DNS
           Evolution . . . . . . . . . . . . . . . . . . . . . . . .  14
       3.13.1.  The Extensions . . . . . . . . . . . . . . . . . . .  15
       3.13.2.  Extensions and Deployment Pressures -- The TXT
                RRTYPE . . . . . . . . . . . . . . . . . . . . . . .  15
       3.13.3.  Periods and Zone Cut Issues  . . . . . . . . . . . .  16
     3.14. Scaling of Reputation and Other Ancillary Information . .  17
     3.15. Tensions among Transport, Scaling, and Content  . . . . .  18
   4.  The Inverse Lookup Requirement  . . . . . . . . . . . . . . .  19
   5.  Internet Scale, Function Support, and Incremental Deployment   20
   6.  Searching and the DNS -- An Historical Note . . . . . . . . .  20
   7.  Security Considerations . . . . . . . . . . . . . . . . . . .  21
   8.  References  . . . . . . . . . . . . . . . . . . . . . . . . .  22
     8.1.  Normative References  . . . . . . . . . . . . . . . . . .  22
     8.2.  Informative References  . . . . . . . . . . . . . . . . .  22
   Acknowledgements  . . . . . . . . . . . . . . . . . . . . . . . .  29
   Author's Address  . . . . . . . . . . . . . . . . . . . . . . . .  29













Klensin                       Informational                     [Page 3]

RFC 8324                      DNS Revisions                February 2018


1.  Introduction



   This document explores contemporary expectations of the Internet's
   domain system (DNS) and compares them to the assumptions and
   properties of the DNS design, including both those documented in the
   RFC Series, an important early paper by the principal author of the
   original RFCs [Mockapetris-1988], and a certain amount of oral
   tradition.  It is primarily intended to ask the question of whether
   the differences are causing enough stresses on the system, stresses
   that cannot be resolved satisfactorily by further patching, that the
   Internet community should be considering designing a new system, one
   that is better adapted to current needs and expectations, and
   developing a deployment and transition strategy for it.  For those
   (perhaps the majority of us) for whom actually replacing the DNS is
   too radical to be realistic, the document may be useful in two other
   ways.  It may provide a foundation for discussing what functions the
   DNS should not be expected to support and how those functions can be
   supported in other ways, perhaps via an intermediate system that then
   calls on the DNS or by using some other type of database technology
   for some set of functions while leaving the basic DNS functions
   intact.  Or it may provide a basis for "better just get used to that
   and the way it works" discussions to replace fantasies about what the
   DNS might do in some alternate reality.

   There is a key design or philosophical question associated with the
   analysis in this document that the document does not address.  It is
   whether changes to perceived requirements to DNS functionality as
   described here are, in most respects, evolutionary or whether many of
   them are instances of trying to utilize the DNS for new requirements
   because it exists and is already deployed independent of whether the
   DNS is really appropriate or not.  The latter might be an instance of
   a problem often described in the IETF as "when all you have is a
   hammer, everything looks like a nail".

   Other recent work, including a short article by Vint Cerf [Cerf2017],
   has discussed an overlapping set of considerations from a different
   perspective, reinforcing the view that it may be time to ask
   fundamental questions about the evolution and future of the DNS.

   While this document does not assume deep technical or operational
   knowledge of the DNS, it does assume some knowledge and at least
   general familiarity with the concepts of RFC 1034 [RFC1034] and RFC
   1035 [RFC1035] and the terminology discussed in RFC 7719 [RFC7719]
   and elsewhere.  Although some of the comments it contains might be
   taken as hints or examples of different ways to think about the
   design issues, it makes no attempt to explore, much less offer a
   tutorial on, alternate naming systems or database technologies.




Klensin                       Informational                     [Page 4]

RFC 8324                      DNS Revisions                February 2018


   It is perhaps worth noting that, while the perspective is different
   and more than a dozen years have passed, many of the issues discussed
   in this document were analyzed and described (most of them with more
   extensive explanations) in a 2005 US National Research Council report
   [NRC-Signposts].

   Readers should note that several references are to obsolete
   documents.  That was done because they are intended to show the
   documents and dates that introduced particular features or concepts.
   When current versions are intended, they are referenced.

2.  Background and Hypothesis



   The Domain Name System (DNS) [RFC1034] was designed starting in the
   early 1980s [RFC0799] [RFC0881] [RFC0882] [RFC0883] with the main
   goal of replacing the flat, centrally administered, host table system
   [RFC0810] [RFC0952] [RFC0953] with a hierarchical, administratively
   distributed, system.  The DNS design included some features that,
   after initial implementation and deployment, were judged to be
   unworkable and either replaced (e.g., the mail destination (MD) and
   mail forwarder (MF) approach [RFC0882] that were replaced by the MX
   approach [RFC0974]), abandoned (e.g., the mechanism for using email
   local parts as labels described in RFC 1034, Section 3.3), or
   deprecated (e.g., the WKS RR TYPE [RFC1123]).  Newer ideas and
   requirements have identified a number of other features, some of
   which were less developed than others.  Of course the original
   designers could not anticipate everything that has come to be
   expected of the DNS in the last 30 years.

   In recent years, demand for new and extended services and uses of the
   DNS have, in turn, led to proposals for DNS extensions or changes of
   various sorts.  Some have been adopted, including a model for
   negotiating extended functionality [RFC2671] (commonly known as
   EDNS(0)) and to support IPv6 [RFC3596], others were found to be
   impracticable, and still others continue to be under consideration.
   Some examples of the latter two categories are discussed below.  A
   few features of the original DNS specification, such as the CLASS
   property and label types, have also been suggested to be so badly
   specified that they should be deprecated [Sullivan-Class].

   Unlike earlier changes such as the Internationalized Domain Names for
   Applications (IDNA) mechanisms for better incorporating non-ASCII
   labels without modifying the DNS structure itself [RFC3490]
   [RFC5890], some recent proposals require or strongly suggest changes
   to APIs, formats, or interfaces by programs that need to retrieve
   information from the DNS or interpret that information.  Differences
   between the DNS architecture and the requirements that imply those
   proposals suggest that it may be time to stop patching the DNS or



Klensin                       Informational                     [Page 5]

RFC 8324                      DNS Revisions                February 2018


   trying to extend it in small increments.  Instead, we should be
   considering moving some current or proposed functionality elsewhere
   or developing a new system that better meets today's needs and a
   transition strategy to it.

   The next section of this document discusses a number of issues with
   the current DNS design that could appropriately be addressed by a
   different and newer design model.  In at least some cases, changing
   the model and protocols could bring significant benefits to the
   Internet and/or its administration.

   This document is not a proposal for a new protocol.  It is intended
   to stimulate thought about how far we want to try to push the
   existing DNS, to examine whether expectations of it are already
   exceeding its plausible capabilities, and to start discussion of a
   redesign or alternatives to one if the time for that decision has
   come.

3.  Warts and Tensions with the Current DNS



   As suggested above, there are many signs that the DNS is incapable of
   meeting contemporary expectations of how it should work and
   functionality it should support.  Some of those expectations are
   unrealistic under any imaginable circumstances; others are impossible
   (or merely problematic) in the current DNS structure but could be
   accommodated in a redesign.  These are examples, rather than a
   comprehensive list, and do not appear in any particular order.

3.1.  Multi-type Queries



   The DNS does not gracefully support multi-type queries.  The current
   case where this problem rears its head involves attempts at solutions
   that return both TYPE A (IPv4) and type AAA (IPv6) addresses
   collectively.  The problem was originally seen with "QTYPE=MAILA"
   [RFC0882] for the original MA and MD RRTYPEs, an experience that
   strongly suggests that some very careful thinking about cache effects
   (and possibly additional DNS changes) would be needed.  Other
   solutions might seem equally or more plausible.  What they, including
   "two types of addresses", probably have in common is that they
   illustrate stresses on the system and that changing the DNS to deal
   with those stresses is not straightforward or likely to be problem-
   free.









Klensin                       Informational                     [Page 6]

RFC 8324                      DNS Revisions                February 2018


3.2.  Matching Part I: Case Sensitivity in Labels and Other Anomalies



   The DNS specifications assume that labels are octet strings and
   octets with the high bit zero have seven-bit ASCII codes in the
   remaining bits.  They require that, when a domain name used in a
   query is matched to one stored in the database, those ASCII
   characters be interpreted in a case-independent way, i.e., upper- and
   lower-case letters are treated as equivalent (digits and symbols are
   not affected) [RFC4343].  For non-ASCII octets, i.e., octets in
   labels with the first bit turned on, there are no assumptions about
   the character coding used, much less any rules about character case
   equivalence -- strings must be compared by matching bits in sequence.
   Even though the current model for handling non-ASCII (i.e.,
   "internationalized") domain name labels (IDNs) [RFC5890] (see
   Section 3.3 below) encodes information so the DNS is not directly
   affected, the notion that some characters in labels are handled in a
   case-insensitive way and that others are case sensitive (or that
   upper case must be prohibited entirely as IDNA does) has caused a
   good deal of confusion and resentment.  Those concerns and complaints
   about inconsistent behavior and mishandling (or suboptimal handling)
   of case relationships for some languages have not been mitigated by
   repeated explanations that the relationships between "decorated"
   lower-case characters and their upper-case equivalents are often
   sensitive to language and locality and therefore not deterministic
   with information available to DNS servers.

3.3.  Matching Part II: Non-ASCII ("Internationalized") Domain Name
      Labels



   Quite independent of the case-sensitivity problem, one of the
   fundamental properties of Unicode [Unicode] is that some abstract
   characters can be represented in multiple ways, such as by a single,
   precomposed, code point or by a base code point followed by one or
   more code points that specify combining characters.  While Unicode
   Normalization can be used to eliminate many (but not all) of those
   distinctions for comparison (matching) purposes, it is best applied
   during matching rather than by changing one string into another.  The
   first version of IDNA ("IDNA2003") made the choice to change strings
   during processing for either storage or retrieval [RFC3490]
   [RFC3491]; the second ("IDNA2008") required that all strings be
   normalized and that upper-case characters are not allowed at all
   [RFC5891].  Neither is optimal, if only because, independent of where
   they are changed if they are changed at all, transforming the strings
   themselves implies that the input string in an application may not be
   the same as the string used in processing and perhaps later display.






Klensin                       Informational                     [Page 7]

RFC 8324                      DNS Revisions                February 2018


   It would almost certainly be preferable, and more consistent with
   Unicode recommendations, to use normalization (and perhaps other
   techniques if they are appropriate) at matching time rather than
   altering the strings at all, even if there were still only a single
   matching algorithm, i.e., normalization were added to the existing
   ASCII-only case folding.  However, even Unicode's discussion of
   normalization [Unicode-UAX15] indicates that there are special,
   language-dependent, cases (the most commonly cited example is the
   dotless "i" (U+0131)).  Not only does the DNS lack any information
   about languages that could be used in a mapping algorithm, but, as
   long as there is a requirement that there be only one mapping
   algorithm for the entire system, that information could not be used
   even if it were available.  One could imagine a successor system that
   would use information stored at nodes in the hierarchy to specify
   different matching rules for subsidiary nodes (or equivalent
   arrangements for non-hierarchical systems).  It is not clear whether
   that would be a good idea, but it certainly is not possible with the
   DNS as we know it.

3.4.  Matching Part III: Label Synonyms, Equivalent Names, and Variants



   As the initial phases of work on IDNs started to conclude, it became
   obvious that the nature and evolution of human language and writing
   systems required treating some names as "the same as" others.  The
   first important example of this involved the relatively recent effort
   to simplify the Chinese writing system, thereby creating a
   distinction between "Simplified" and "Traditional" Chinese even
   though the meaning of the characters remained the same in almost all
   cases (in so-called ideographic character sets, characters have
   meaning rather than exclusively representing sounds).  A joint effort
   among the relevant Country Code Top-Level Domain (ccTLD) registries
   and some other interested parties produced a set of recommendations
   for dealing with the issues with that script [RFC3743] and introduced
   the concept of "variant" characters and domain names.

   However, when names are seen as having meanings, rather than merely
   being mnemonics, especially when they represent brands or the
   equivalent, or when spelling for a particular written language is not
   completely standardized, demands to treat different strings as exact
   equivalents are obvious and inevitable.  As a trivial English-
   language example, it is widely understood that "colour" and "color"
   represent the same word, so does that imply that, if they are used as
   DNS labels in domain names all of whose other labels are identical,
   the two domain names should be treated as identical?  Examples for
   other languages or writing systems, especially ones in which some or
   all markings that distinguish characters or words by sound or tone or
   that change the pronunciation of words are optional, are often more
   numerous and more problematic than national spelling differences in



Klensin                       Informational                     [Page 8]

RFC 8324                      DNS Revisions                February 2018


   English, but they are harder to explain to those unfamiliar with
   those other languages or writing systems (and hard to illustrate in
   ASCII-only Internet-Drafts and RFCs).  Although approximations are
   possible, the DNS cannot handle that requirement: not only do its
   aliasing mechanisms (CNAME, DNAME, and various proposals for newer
   and different types of aliasing [DNS-Aliases] [DNS-BNAME]) not
   provide a strong enough binding, but the ability to use those aliases
   from a subtree controlled by one administrative entity to that of
   another one implies that there is little or no possibility of the
   owner (in either the DNS sense or the registrar-registrant one) of a
   particular name to control the synonyms for it.  Some of that issue
   can be dealt with at the application level, e.g., by redirects in web
   protocols, but taking that approach, which is the essential
   characteristic of "if both names belong to the same owner, everything
   is OK" approaches, results in names being handled in inconsistent
   ways in different protocols.

   A different way of looking at part of this issue (and, to some
   degree, of the one discussed above in Section 3.3) is that these
   perceived equivalences and desired transformations are context-
   dependent, but the DNS resolution process is not [RFC6912].

   Similar problems arise as people notice that some characters are
   easily mistaken for others and that might be an opportunity for user
   confusion and attacks.  Commonly cited examples include the Latin and
   Cyrillic script "a" characters, which are identical [CACM-Homograph],
   the characters in many scripts that look like open circles or
   vertical or horizontal lines, and even the Latin script letter "l"
   and the European digit "1", but examples abound in other scripts and
   combinations of scripts as well.  The most common proposed solution
   within the DNS context has been to treat these cases, as well as
   those involving orthographic variations, as "variants" (but variants
   different from the system for Chinese characters mentioned above) and
   either ban all but one (or a few) of the possible labels from the DNS
   (possibly on a first come, first served basis) or ensure that any
   collection of such strings that are delegated as assigned to the same
   ownership (see above).  Neither solution is completely satisfactory:
   if all but one string is excluded, users who guess at a different
   form, perhaps in trying to transcribe characters from written or
   printed form, don't find what they are looking for and, as pointed
   out above, "same ownership" is sufficient only with carefully
   designed and administered applications protocol support, and
   sometimes not then.

   Some of these issues are discussed at more length in an ICANN report
   [ICANN-VIP].





Klensin                       Informational                     [Page 9]

RFC 8324                      DNS Revisions                February 2018


3.5.  Query Privacy



   There has been growing concern in recent years that DNS queries occur
   in cleartext on the public Internet and that, if those queries can be
   intercepted, they can expose a good deal of information about
   interests and contacts that could compromise individual privacy.
   While a number of proposals, including query name minimization
   [RFC7816] and running DNS over an encrypted tunnel [RFC7858], have
   been made to mitigate that problem, they all appear to share the
   common properties of security patches rather than designed-in
   security or privacy mechanisms.  While experience may prove otherwise
   once (and if) they are widely deployed, it does not appear that any
   of them are as satisfactory as a system with query privacy designed
   in might be.  More general tutorials on this issue have appeared
   recently [Huston2017a].

3.6.  Alternate Namespaces for Public Use in the DNS Framework: The
      CLASS Problem



   The DNS standards include specification of a CLASS value, which
   "identifies a protocol family or instance of a protocol" (RFC 1034,
   Section 3.6, and elsewhere).  While CLASS was used effectively in the
   early days of the DNS to manage different protocol families within
   the same administrative environment, recent attempts to use it to
   either partition the DNS namespace in other ways such as for
   non-ASCII names (partially to address the issues in Sections 3.2 and
   3.3) or use DNS mechanisms for entirely different namespaces have
   exposed fundamental problems with the mechanism [Sullivan-Class].
   Perhaps the most fundamental of those problems is disagreement about
   whether multiple CLASSes were intended to exist within a given zone
   (with records within RRSETs) or whether different CLASSes implied
   different zones.  Different implementations make different
   assumptions [Faltstrom-2004] [Vixie-20170704].  These problems have
   led to recommendations that it be dropped entirely [Sullivan-Class],
   but discussions on the IETF list and in WGs in mid-2017 made it clear
   that there is no clear consensus on that matter.

3.7.  Loose Synchronization



   The DNS model of master and slave servers, with the latter initiating
   updates based on expiration interval values, and local caches with
   updates based on TTL values, depends heavily on an approach that has
   come to be called "loose synchronization", i.e., that there can be no
   expectation that all of the servers that might reasonably answer a
   query will have exactly the same data unless those data have been
   unchanged for a rather long period.  Put differently, if some or all
   of the records associated with a particular node in the DNS




Klensin                       Informational                    [Page 10]

RFC 8324                      DNS Revisions                February 2018


   (informally, a fully qualified domain name (FQDN)) change, one cannot
   expect those changes to be propagated immediately.

   That model has worked rather well since the DNS was first deployed,
   protecting the system from requirements for mechanisms that are
   typical where a simultaneous update of multiple systems is needed.
   Such mechanisms include elaborate locking, complex update procedures
   and handshaking, or journaling.  As has often been pointed out with
   the Internet, implementation and operational complexity are often the
   enemy of stability, security, and robustness.  Loose synchronization
   has helped keep the DNS as simple and robust as possible.

   A number of recent ideas about using the DNS to store data for which
   important changes occur very rapidly are, however, largely
   incompatible with loose synchronization.  Efforts to use very short
   (or zero) refresh times (in SOA records for slave updates from
   masters) and TTLs (for caches) to simulate nearly simultaneous
   updating may work up to a point but appear to impose very heavy loads
   on servers and distribution mechanisms that were not designed to
   accommodate that style of working.  Similar observations can be made
   about attempts to use the NOTIFY extension [RFC1996] or dynamic,
   "server-push", updating rather than the traditional DNS mechanisms.
   While the NOTIFY and push mechanisms normally provide refresh times
   and update mechanisms faster than those specified in RFCs 1034 and
   1035, they imply that a "master" server must know the identities of
   (and have good connectivity to all of) its slaves.  That defeats at
   least some of the advantages associated with stealth slaves,
   particularly those associated with reduction of query traffic across
   the Internet.  Those mechanisms do nothing for cache updates: unless
   servers keep track of the source of every query for names associated
   with a specific zone and then somehow notify the query source
   systems, the only alternative to having information that might be
   obsolete stored in caches is to use very short or zero TTLs so the
   cached data time out almost immediately after being stored (or are
   not stored at all), requiring a new query to an authoritative server
   each time a resolver attempts to look up a name.

3.8.  Private Namespaces and Special Names



   Almost since the DNS was first deployed, there have been situations
   in which it is desirable to use DNS-like names, and often DNS
   resolution mechanisms or modifications of them, with namespaces for
   which globally available and consistent resolution using the public
   DNS is either unfeasible or undesirable (and for which the use of
   CLASS is not an appropriate mechanism).  The need to isolate names
   and addresses on LANs from the public Internet, typically via "split
   horizon" approaches, is one example of this requirement although
   often not recognized as such.  Another example that has generated a



Klensin                       Informational                    [Page 11]

RFC 8324                      DNS Revisions                February 2018


   good deal of controversy involves "special names" -- labels or
   pseudo-labels, often in TLD positions, that signal that the full name
   should not be subject to normal DNS resolution or other processing
   [RFC6761] [RFC8244].

   Independent of troublesome policy questions about who should allocate
   such names and the procedures to be used, they almost inherently
   require either a syntax convention to identify them (there actually
   was such a convention, but it was abandoned many years ago and there
   is no plausible way to reinstitute it) or tables of such names that
   are known to, and kept updated on, every resolver on the Internet, at
   least if spurious queries to the root servers are to be avoided.

   If the DNS were to be redesigned and replaced, we could recognize
   this requirement as part of the design and handle it much better than
   it is possible to handle it today.

3.9.  Alternate Query or Response Encodings



   The DNS specifies formats for queries and data responses, based on
   the state of the art and best practices at the time it was designed.
   Recent work has suggested that there would be significant advantages
   to supporting at least a description of the DNS messages in one or
   more alternate formats, such as JSON [Hoffman-DNS-JSON]
   [Hoffman-SimpleDNS-JSON].  While that work has been carefully done to
   avoid requiring changes to the DNS, much of the argument for having
   such a JSON-based description format could easily be turned into an
   argument that, if the DNS were being revised, that format might be
   preferable as a more direct alternative to having DNS queries and
   responses in the original form.

3.10.  Distribution and Management of Root Servers



   The DNS model requires a collection of root servers that hold, at
   minimum, information about top-level domains.  Over the years, that
   requirement has evolved from a technically fairly minor function,
   normally carried out as a service to the broader Internet community
   and its users and systems, to a subject that is intensely
   controversial with regard to control of those servers, including how
   they should be distributed and where they should be located.  While a
   number of mechanisms, most recently including making the information
   more local [RFC7706], have been proposed and one (anycast [RFC7094])
   is in very active use to mitigate some of the real and perceived
   problems, it seems obvious that a DNS successor, designed for today's
   global Internet and perceived requirements, could handle these
   problems in a technically more appropriate and less controversial
   way.  Some additional discussion of the issues involved appears in a
   recent paper [Huston2017b].



Klensin                       Informational                    [Page 12]

RFC 8324                      DNS Revisions                February 2018


3.11.  Identifiers versus Brands and Other Convenience Names



   A key design element of the original network object naming systems
   for the ARPANET, largely inherited by the DNS, was that the names,
   while expected to be mnemonic, were identifiers and their being
   highly distinguishable and not prone to ambiguity was important.
   That led to restrictive rules about what could appear in a name.
   Those restrictions originated with the host table and even earlier
   [RFC0236] [RFC0247] and came to the DNS (largely via SMTP) as the
   "preferred syntax" (RFC 1034, Section 3.5) or what we now often call
   the letter-digit-hyphen (LDH) rule.  Similar rules to make
   identifiers easier to use, less prone to ambiguity, or less likely to
   interfere with syntax occur frequently in more formal languages.  For
   example, almost every programming language has restrictions on what
   can appear in an identifier, and Unicode provides general
   recommendations about identifier composition [Unicode-USA31].  Both
   are quite restrictive as compared to the number of characters and
   total number of strings that can be written using that character
   coding system.

   That model, which originally prohibited labels starting with digits
   in order to avoid any possible confusion with IP addresses, began to
   break down in 1987 or 1988 when a company named 3Com wanted to use
   its corporate name as a label within the COM TLD, and the rule was
   relaxed [RFC1123].

   In the last decade or two, the perspective that company names should
   be supported if possible has expanded and done so largely without its
   limits, if any, being explicitly understood or acknowledged.  In the
   current form, the DNS is really (and primarily) a system for
   expressing thoughts and concepts.  Those include free expression of
   ideas in as close to natural language as possible as well as
   representation of product names and brands.  That view requires
   letter-like characters that might not be reasonable in identifiers
   along with a variety of symbols and punctuation.  It may also require
   indicators of preferred type styles to provide information in a form
   that exactly matches personal or legal preferences.  At least if
   carried to an extreme, that perspective would argue for standardizing
   word and sentence separators, removing the limit of 63 octets per
   label and probably the limit of 255 octets on the total length of a
   domain name, and perhaps even eliminating the hierarchy or allowing
   separators for labels in presentation form (now fixed at "." for the
   DNS) to be different according to context.  It suggests that, at
   least, the original design was defective in not prioritizing those
   uses over the more restrictive approach associated with prioritizing
   unique and unambiguous identifiers.





Klensin                       Informational                    [Page 13]

RFC 8324                      DNS Revisions                February 2018


   So we have two or, depending on how one counts, three very different
   use cases.  The historical one is support for unique identifiers.
   The other is expression of ideas and, if one considers them separate,
   presentation of brand and product names.  Because they inherently
   involve different constraints, priorities, and success criteria,
   these perspectives are, at best, only loosely compatible.

   We cannot simultaneously optimize both the identifier perspective and
   either or both of the others in the same system.  At best, there are
   some complex trade-offs involved.  Even then, it is not clear that
   the same DNS (or other system) can accommodate all of them.  Until we
   come to terms with that, the differences manifest themselves with
   friction among communities, most often with tension between "we want
   to do (or use or sell) these types of labels" and "not good for the
   operational Internet or the DNS".

3.12.  A Single Hierarchy with a Centrally Controlled Root



   A good many Internet policy discussions in the last two decades have
   revolved around such questions of how many top-level domains there
   should be, what they should be, who should control them and how, how
   (or if) their individual operations and policy decisions should be
   accountable to others, and what processes should be used (and by what
   entities or organizational structures) to make those decisions.
   Several people have pointed out that, if we were designing a next-
   generation DNS using today's technology, it should be possible to
   remove the technical requirement for a central authority over the
   root (some people have suggested that blockchain approaches would be
   helpful for this purpose; others believe they just would not scale
   adequately, at least at acceptable cost, but that other options are
   possible).  Whether elimination of a single, centrally controlled,
   root would be desirable or not is fairly obviously a question of
   perspective and priorities.

3.13.  Newer Application Protocols, New Requirements, and DNS Evolution



   New work done in other areas has led to demands for new DNS features,
   many of them involving data values that require recursively
   referencing the DNS.  Early record types that did that were
   accompanied by restrictions that reduced the risk of looping
   references or other difficulties.  For example, while the MX RRTYPE
   has a fully qualified domain name as its data, SMTP imposes "primary
   name" restrictions that prevent the name used from being, e.g., a
   CNAME.  While loops with CNAMEs are possible, Section 3.6 of RFC 1034
   includes a discussion about ways to avoid problems and how they
   should be handled.  Some newer protocols and conventions can cause
   more stress.  There are separate issues with additions and with how
   the DNS has been extended to try to deal with them.



Klensin                       Informational                    [Page 14]

RFC 8324                      DNS Revisions                February 2018


3.13.1.  The Extensions



   Some examples of DNS extensions for new protocol demands that
   illustrate, or have led to, increased stress include:

   NAPTR:  Requires far more complex data in the DNS for ENUM (e.g.,
      Voice over IP (VoIP), specifically SIP) support, including URI
      information and hence recursive or repeated lookups, than any of
      the RRTYPEs originally supported.  The RRSET associated with these
      records can become quite large because the separator between the
      various records is part of the RDATA, and not the {owner, class,
      type} triple (a problem slightly related to the problem with
      overloading of TXT RRTYPE discussed in Section 3.13.2).  This
      problem, and similar ones for some of the cases below. may suggest
      that any future design is in need of a different TYPE model such
      as systematic arrangements for subtypes or some explicit hierarchy
      in the TYPEs.

   URI:  Has a URI as its data, typically also requiring recursive or
      repeated lookups.

   Service location (SRV) and credential information (including Sender
   Policy Framework (SPF) and DomainKeys Identified Mail (DKIM)):
      Require structured data and, especially for the latter two,
      significantly more data than most original RRTYPEs.

   URI/URL:  The early design decision for the World Wide Web that its
      mechanism for identifying digital web content (now known as
      Uniform Resource Identifiers [RFC3986]) did so by using domain
      names and hence the network location of the information or other
      material.  That, in turn, has required systems intended to improve
      web performance by locating and retrieving a "nearest copy"
      (rather than the single copy designated by the URL) to intercept
      DNS queries and respond with values that are not precisely those
      stored for the designated domain name in the DNS or to otherwise
      access information in a way not supported by the DNS itself.

3.13.2.  Extensions and Deployment Pressures -- The TXT RRTYPE



   Unfortunately (but unsurprisingly), and despite IETF efforts to make
   things easier [RFC6895], DNS support libraries have often been slow
   to add full support for new RRTYPEs.  This has impeded deployment of
   applications that depend on those types and that must ask (query)
   explicitly for them.  Both to get faster deployment and, at least
   until recently, to avoid burdensome IETF approval procedures, many
   application designers have chosen to push protocol-critical





Klensin                       Informational                    [Page 15]

RFC 8324                      DNS Revisions                February 2018


   information into records with TXT RRTYPE, a record type that was
   originally intended to include only information equivalent to
   comments.

   This causes two problems.  First, TXT records used this way tend to
   get long and complex, which is a problem in itself if one is trying
   to minimize TCP connections.  Second, applications that are
   attempting to obtain data cannot merely ask for the relevant QTYPE;
   they must obtain all of the records with QTYPE TXT and parse them to
   determine which ones are of interest.  That would be easier if there
   was some standard for how to do that parsing, but, at least in part
   because the clear preference in the DNS design is for distinct
   RRTYPEs for different kinds of information, there is no such
   standard.  (There was a proposal in 1993 to structure the TXT DATA in
   a way that would have addressed the issue [RFC1464], but it
   apparently never went anywhere.)

   On the other hand, this issue is somewhat different from most of the
   others described in this document because (as the IETF has
   recommended several times) the problem is easily solved within the
   current DNS design by allocating and supporting new RRTYPEs when
   needed rather than using TXT as a workaround (that does not mean that
   other solutions are impossible, either with the current DNS or with
   some other design).  The problem then lies in the implementations
   and/or mechanisms that deter or impede rapid deployment of support
   for new RRTYPEs.

3.13.3.  Periods and Zone Cut Issues



   One of the DNS characteristics that is poorly understood by
   non-experts is that the period (".", U+002E) character can be used in
   four different ways:

   o  As a label separator in the presentation form that also designates
      a "zone break" (delegation boundary).  For example,
      foo.bar.example.com indicates the owner, "foo", of records in the
      "bar.example.com" zone.

   o  As a label separator in the presentation form that does not
      designate a zone break.  For example, foo.bar.example.com
      indicates the owner, "foo.bar", of records in the "example.com"
      zone.

   o  As a character within a label, including as a substitute for an
      at-sign ("@") when an email address appears in an SOA record or in
      a label that denotes such an address (see Section 2 above).  The
      ability to embed periods in labels in this way has also led to
      attacks in which, e.g., a domain name consisting of the labels



Klensin                       Informational                    [Page 16]

RFC 8324                      DNS Revisions                February 2018


      "example" followed by "com" is deliberately confused with the
      single label "example.com" with an embedded period.

   o  At the end of a fully qualified domain name to designate the root
      zone, e.g., "example.com." (RFC 1034, Section 3.1).

   In general, these cases cannot be distinguished by looking at them.
   The third is problematic for non-DNS reasons, e.g.,
   "john.doe.example.net" can be interpreted as either a simple FQDN or
   as a notation for john@doe.example.net, john.doe@example.net, or even
   (at least in principle) john.doe.example@net.

   The distinction between the FQDN interpretation and the first email-
   like one was probably not important as the DNS was originally
   intended to be used.  However, as soon as RRTYPEs (other than NS
   records that define the zone cut) are used that are sensitive to the
   boundaries between zones, the distinctions become important to people
   other than the relevant zone administrators.  DNSSEC [RFC4033]
   involves one such set of relationships.  It increases the importance
   of questions about what should go in a parent zone and what should go
   in child zones and how much difference it makes if NS records in a
   parent zone for a child zone are consistent with the records and data
   in the child zone.  This also causes application issues and may raise
   questions about relationships between registrars and one or more
   registries or, if they are separate, DNS operators.

3.14.  Scaling of Reputation and Other Ancillary Information



   The original design for DNS administration, reflected in RFC 1591
   [RFC1591] and elsewhere, assumed that all domains would exhibit a
   very high level of responsibility toward and for the community and
   that level of responsibility would be enforced if necessary.

   More recent decisions, many of them associated with commercialization
   of the DNS, have eroded those very strong assumptions of registry
   responsibility and accountability to the point that many consider
   decisions about delegation of names, identification of registrants,
   and relationships among names to be matters of "registrant beware"
   and even "user and applications beware".  While some recent protocols
   and proposals at least partially reflect that original model of a
   high level of responsibility (see, e.g., IDNA [RFC5890] and a more
   recent discussion [Klensin-5891bis]), other decisions and actions
   tend to shift responsibility to the registrant or try to avoid
   accountability entirely.  One possible approach to the problems,
   especially security problems, that are enabled by those new trends
   and the associated environment is to establish reputation systems
   associated with clearly defined administrative boundaries and with




Klensin                       Informational                    [Page 17]

RFC 8324                      DNS Revisions                February 2018


   warnings to users, even if those reputation systems are managed by
   parties not directly associated with the DNS.

   The IETF DBOUND WG [IETF-DBOUND] addressed ways to establish and
   document boundaries more precise than simple dependencies on TLDs,
   but it was not successful in producing a standard.

   A TLD reputation-based approach was adopted by some web browsers
   after IDNs and a growing number of Generic Top-Level Domains (gTLDs)
   were introduced; that approach was based on a simple list and does
   not scale to the current size of the DNS or even the DNS root.

3.15.  Tensions among Transport, Scaling, and Content



   The original design for the DNS envisaged a simple query and response
   protocol where both the command and the response could be readily
   mapped into a single IP packet.  The host requirements specification
   [RFC1123] required all DNS applications to accept a UDP query or
   response over UDP with up to 512 octets of DNS payload.  TCP was seen
   as a fallback when the response was greater than this 512-octet
   limit, and this fallback to use TCP as the transport protocol was
   considered to be the exception rather than the rule.

   Over the intervening years, we have seen the rise of a common
   assumption of an Internet-wide Maximum Transmission Unit (MTU) size
   of 1,500 octets, accompanied with an assumption that UDP
   fragmentation is generally viable.  This underpins the adoption of
   the Extension Mechanisms for DNS (EDNS(0)) [RFC6891] to, among other
   things, specify a UDP buffer size larger than 512 octets and a
   suggestion within that specification to use 4,096 as a suitable
   compromise for the UDP payload size.  This has proved to be
   fortuitous for the DNSSEC security extensions where the addition of
   DNSSEC security credentials in DNS responses [RFC4034] can lead to
   the use of large DNS responses.  However, this exposes some tensions
   over the handling of fragmentation in IP, where UDP fragments have
   been observed to be filtered by various firewalls.  Additionally for
   IPv6, there are the factors of filtering the ICMPv6 Packet Too Big
   diagnostic messages and discarding the IPv6 packets that contain
   extension headers [RFC7872].  More generally, fragmented UDP packets
   appear to have a lower level of reliability than unfragmented TCP
   packets.

   Behind this observation about relative reliability of delivery is the
   tension between the lightweight load of UDP and the downside of
   elevated probability of discarding of packet fragments as compared to
   TCP, which offers increased levels of assurance of content delivery,
   but with the associated imposition of TCP session state and the
   downside of reduced DNS scalability and increased operational cost.



Klensin                       Informational                    [Page 18]

RFC 8324                      DNS Revisions                February 2018


4.  The Inverse Lookup Requirement



   The requirement for an inverse lookup capability, i.e., the ability
   to find a domain name given an address and, in principle, to find the
   owner of a record by any of its data elements, was recognized in RFC
   882.  The feature was identified as optional but carried forward into
   RFCs 1034 and 1035 but was explicitly deprecated by RFC 1034 for
   address-to-hostname lookup (although RFC 1035 uses exactly that type
   of lookup in an example).  Despite the discussion of inverted forms
   of the database in RFC 1035, inverse lookup has rarely, if ever, been
   implemented, at least in its general form.  The fundamental
   difficulties with inverse lookup in either the form described in RFC
   882 or the "in-addr.arpa" approach mentioned below are consistent
   with the problems described in fundamental papers on database
   management [Codd1970] but were not described in RFC 1035 or related
   contemporary IETF documents.

   It is interesting to speculate on how many of the current
   requirements to treat aliases as an integrated set of synonyms (e.g.,
   for variant handling) would have been addressed if inverse lookups
   could reliably produce the owners of CNAME records.

   At the same time, it was obviously important to have some mechanism
   for address-to-name resolution.  It was provided by PTR RRTYPE
   entries in the IN-ADDR.ARPA zone, with delegations on octet
   boundaries.  However, that approach required that information be
   maintained in parallel, in separate zones, for the name-to-address
   and address-to-name mappings.  That synchronization requirement for
   two copies of essentially the same data was another popular topic in
   the database management literature a decade or more before the DNS
   and, predictably, led to many inconsistencies and other failures.

   The introduction of Classless Inter-Domain Routing (CIDR) [RFC1518]
   and Provider-Dependent addresses made the situation even more
   difficult, because it was no longer possible to delegate the
   administration of reverse mapping records for small networks to the
   actual operators of those networks.  ISPs and other aggregators often
   had no incentive to maintain reverse mapping records consistent with
   network operator assignment of domain names.  A proposal to use
   binary labels to work around that issue [RFC2673] was abandoned
   somewhat over three years later [RFC6891].

   Independent of how much or little harm the absence of a general
   inverse lookup facility has caused and how effective the
   "in-addr.arpa" approach has been, inverse lookup remains a facility
   that was anticipated and known to be useful in the original DNS
   design but that has never been fully realized.




Klensin                       Informational                    [Page 19]

RFC 8324                      DNS Revisions                February 2018


5.  Internet Scale, Function Support, and Incremental Deployment



   In addition to the stresses caused by the new functions, including
   those described in Section 3.13, incremental deployment of systems
   that utilize them means that some functions will work in some
   environments and not others.  This has been especially problematic
   with complex, multi-record, capabilities like DNSSEC that provide or
   require special validation mechanisms and with some EDNS(0)
   extensions [RFC6891] that require both the client and server to
   accept particular extensions.  When DNS functionality is required in
   embedded devices, deployment of new features across the entire
   Internet in a reasonable period of time is nearly impossible.

   If one were redesigning the DNS, one could imagine ways to address
   these issues that would make them slightly more tractable, and, of
   course, the features that are known to be necessary today could
   become part of the baseline, "mandatory to implement", specification.

6.  Searching and the DNS -- An Historical Note



   Some of the issues identified above might reasonably be addressed,
   not by changing the DNS itself but by changing our model of what it
   is about and how it is used.  Specifically, one key assumption when
   the DNS (and the host table system before it) was designed was that
   it was a naming system for network resources, not, e.g., digital
   content.  As such, exact matching was important, it was reasonable to
   have labels treated as mnemonics that did not necessarily have
   linguistic or semantic meaning except to those using them, and so on.
   A return to that model, presumably by having user-facing applications
   call on an intermediate layer to disambiguate user-friendly names and
   map them to DNS names (or network object locators more generally),
   would significantly reduce stress on the DNS and would also allow
   dealing with types of matching and similar or synonymous strings that
   cannot be handled algorithmically no matter how much DNS matching
   rules were altered.

   In some respects, search engines based on free-text analysis and
   linkages among information have come to serve many of the functions
   of such an intermediate layer.  Many studies and sources have pointed
   out that few users actually understand, much less care about, the
   distinction between a DNS name and a search term.  Recent versions of
   some web browsers have both recognized the failure of that
   distinction and reinforced it by eliminating the separation between
   "URL" and "search bar".







Klensin                       Informational                    [Page 20]

RFC 8324                      DNS Revisions                February 2018


   It is worth noting that, while that "search" approach, or some other
   approach that abstracted and separated several of the issues
   identified in Section 3 from the DNS protocol and database
   themselves, it does not address all of them.  At least some elements
   of several of those issues, such as the synchronization ones
   described in Section 3.7 and the transport ones described in
   Section 3.15, are inherent in the DNS design, and, if we are not
   going to replace the DNS, we had best get used to them.

   In the early part of the last decade, the IETF engaged in some
   preliminary exploration of the intermediate-layer approach in the
   context of IDNs and what were then called "Internet keywords"
   [DNS-search].  While that exploratory effort met several times
   informally, it never became an organized IETF activity, largely
   because of the choice of what became the IDNA approach but also in
   part by signs that the "Internet keywords" efforts were beginning to
   fall apart.

   It may be time to reexamine intermediate-layer approaches.  If so,
   the effort should examine use of those approaches by appropriate
   user-facing applications that might be used to address some of the
   issues identified above.  The Internet and the DNS have changed
   considerably since the 2000-2003 period.  Several of those changes
   are discussed elsewhere in this document; others, including
   repurposing of the DNAME RRTYPE from support for transitions
   [RFC2672] to a general-purpose mechanism for aliases of subtrees
   [RFC6672] and the addition of over a thousand new TLDs
   [IANA-TLD-registry], are not but nonetheless are part of the context
   for intermediate-layer work that did not exist in 2003.

7.  Security Considerations



   A wide range of security issues related to both securing the DNS and
   also to abilities to use namespaces for nefarious purposes have
   arisen.  Issues of securing the DNS would obviously be essential to a
   replacement of the DNS.  Issues of preventing nefarious use of the
   namespace (e.g. use of the name that appears or disappears as a
   signal to bots) would appear to be harder to solve within the naming
   system.












Klensin                       Informational                    [Page 21]

RFC 8324                      DNS Revisions                February 2018


8.  References



8.1.  Normative References



   [RFC1034]  Mockapetris, P., "Domain names - concepts and facilities",
              STD 13, RFC 1034, DOI 10.17487/RFC1034, November 1987,
              <https://www.rfc-editor.org/info/rfc1034>.

   [RFC1035]  Mockapetris, P., "Domain names - implementation and
              specification", STD 13, RFC 1035, DOI 10.17487/RFC1035,
              November 1987, <https://www.rfc-editor.org/info/rfc1035>.

8.2.  Informative References



   [CACM-Homograph]
              Gabrilovich, E. and A. Gontmakher, "The Homograph Attack",
              Communications of the ACM, Volume 45, Issue 2, pp. 128,
              DOI 10.1145/503124.503156, February 2002,
              <http://www.cs.technion.ac.il/~gabr/papers/
              homograph_full.pdf>.

   [Cerf2017] Cerf, V., "Desirable Properties of Internet Identifiers",
              IEEE Internet Computing, Volume 21, Issue 6, pp. 63-64,
              DOI 10.1109/MIC.2017.4180839, November/December 2017.

   [Codd1970] Codd, E., "A Relational Model of Data for Large Shared
              Data Banks", Communications of the ACM, Volume 13, Issue
              6, pp. 377-387, DOI 10.1145/362384.362685, June 1970,
              <https://dl.acm.org/citation.cfm?id=362685>.

   [DNS-Aliases]
              Woolf, S., Lee, X., and J. Yao, "Problem Statement: DNS
              Resolution of Aliased Names", Work in Progress,
              draft-ietf-dnsext-aliasing-requirements-01, March 2011.

   [DNS-BNAME]
              Yao, J., Lee, X., and P. Vixie, "Bundled DNS Name
              Redirection", Work in Progress, draft-yao-dnsext-bname-06,
              May 2016.

   [DNS-search]
              IETF, "Internet Resource Name Search Service (IRNSS)",
              2003, <https://datatracker.ietf.org/wg/irnss/about/>.

   [Faltstrom-2004]
              Faltstrom, P. and R. Austein, "Design Choices When
              Expanding DNS", Work in Progress,
              draft-ymbk-dns-choices-00, May 2004.



Klensin                       Informational                    [Page 22]

RFC 8324                      DNS Revisions                February 2018


   [Hoffman-DNS-JSON]
              Hoffman, P., "Representing DNS Messages in JSON", Work in
              Progress, draft-hoffman-dns-in-json-13, October 2017.

   [Hoffman-SimpleDNS-JSON]
              Hoffman, P., "Simple DNS Queries and Responses in JSON",
              Work in Progress, draft-hoffman-simplednsjson-01, November
              2017.

   [Huston2017a]
              Huston, G. and J. Silva Dama, "DNS Privacy", The Internet
              Protocol Journal, Vol. 20, No. 1, March 2017,
              <http://ipj.dreamhosters.com/wp-content/uploads/
              issues/2017/ipj20-1.pdf>.

   [Huston2017b]
              Huston, G., "The Root of the Domain Name System", The
              Internet Protocol Journal, Vol. 20, No. 2, pp. 15-25, June
              2017, <http://ipj.dreamhosters.com/wp-content/uploads/
              2017/08/ipj20-2.pdf>.

   [IANA-TLD-registry]
              Internet Assigned Numbers Authority (IANA), "Root Zone
              Database", <https://www.iana.org/domains/root/db>.

   [ICANN-VIP]
              ICANN, "IDN Variant Issues Project: Final Integrated
              Issues Report Published and Proposed Project Plan for Next
              Steps is Now Open for Public Comment", February 2012,
              <https://www.icann.org/news/announcement-2012-02-20-en>.

   [IETF-DBOUND]
              IETF, "Domain Boundaries (dbound)", 2017,
              <https://datatracker.ietf.org/wg/dbound/about/>.

   [Klensin-5891bis]
              Klensin, J. and A. Freytag, "Internationalized Domain
              Names in Applications (IDNA): Registry Restrictions and
              Recommendations", Work in Progress,
              draft-klensin-idna-rfc5891bis-01, September 2017.











Klensin                       Informational                    [Page 23]

RFC 8324                      DNS Revisions                February 2018


   [Mockapetris-1988]
              Mockapetris, P. and K. Dunlap, "Development of the Domain
              Name System", SIGCOMM '88 Symposium, pp. 123-133,
              available from ISI Reprint Series, ISI/RS-88-219
              <ftp://ftp.isi.edu/isi-pubs/rs-88-219.pdf>,
              DOI 10.1145/52324.52338, August 1988,
              <http://dl.acm.org/citation.cfm?id=52338>.

   [NRC-Signposts]
              National Research Council, Signposts in Cyberspace: The
              Domain Name System and Internet Navigation,
              ISBN 0-309-54979-5, 2005, <https://www.nap.edu/
              catalog/11258/signposts-in-cyberspace-the-domain-name-
              system-and-internet-navigation>.

   [RFC0236]  Postel, J., "Standard host names", RFC 236,
              DOI 10.17487/RFC0236, September 1971,
              <https://www.rfc-editor.org/info/rfc236>.

   [RFC0247]  Karp, P., "Proffered set of standard host names", RFC 247,
              DOI 10.17487/RFC0247, October 1971,
              <https://www.rfc-editor.org/info/rfc247>.

   [RFC0799]  Mills, D., "Internet name domains", RFC 799,
              DOI 10.17487/RFC0799, September 1981,
              <https://www.rfc-editor.org/info/rfc799>.

   [RFC0810]  Feinler, E., Harrenstien, K., Su, Z., and V. White, "DoD
              Internet host table specification", RFC 810,
              DOI 10.17487/RFC0810, March 1982,
              <https://www.rfc-editor.org/info/rfc810>.

   [RFC0881]  Postel, J., "Domain names plan and schedule", RFC 881,
              DOI 10.17487/RFC0881, November 1983,
              <https://www.rfc-editor.org/info/rfc881>.

   [RFC0882]  Mockapetris, P., "Domain names: Concepts and facilities",
              RFC 882, DOI 10.17487/RFC0882, November 1983,
              <https://www.rfc-editor.org/info/rfc882>.

   [RFC0883]  Mockapetris, P., "Domain names: Implementation
              specification", RFC 883, DOI 10.17487/RFC0883, November
              1983, <https://www.rfc-editor.org/info/rfc883>.

   [RFC0952]  Harrenstien, K., Stahl, M., and E. Feinler, "DoD Internet
              host table specification", RFC 952, DOI 10.17487/RFC0952,
              October 1985, <https://www.rfc-editor.org/info/rfc952>.




Klensin                       Informational                    [Page 24]

RFC 8324                      DNS Revisions                February 2018


   [RFC0953]  Harrenstien, K., Stahl, M., and E. Feinler, "Hostname
              Server", RFC 953, DOI 10.17487/RFC0953, October 1985,
              <https://www.rfc-editor.org/info/rfc953>.

   [RFC0974]  Partridge, C., "Mail routing and the domain system",
              STD 10, RFC 974, DOI 10.17487/RFC0974, January 1986,
              <https://www.rfc-editor.org/info/rfc974>.

   [RFC1123]  Braden, R., Ed., "Requirements for Internet Hosts -
              Application and Support", STD 3, RFC 1123,
              DOI 10.17487/RFC1123, October 1989,
              <https://www.rfc-editor.org/info/rfc1123>.

   [RFC1464]  Rosenbaum, R., "Using the Domain Name System To Store
              Arbitrary String Attributes", RFC 1464,
              DOI 10.17487/RFC1464, May 1993,
              <https://www.rfc-editor.org/info/rfc1464>.

   [RFC1518]  Rekhter, Y. and T. Li, "An Architecture for IP Address
              Allocation with CIDR", RFC 1518, DOI 10.17487/RFC1518,
              September 1993, <https://www.rfc-editor.org/info/rfc1518>.

   [RFC1591]  Postel, J., "Domain Name System Structure and Delegation",
              RFC 1591, DOI 10.17487/RFC1591, March 1994,
              <https://www.rfc-editor.org/info/rfc1591>.

   [RFC1996]  Vixie, P., "A Mechanism for Prompt Notification of Zone
              Changes (DNS NOTIFY)", RFC 1996, DOI 10.17487/RFC1996,
              August 1996, <https://www.rfc-editor.org/info/rfc1996>.

   [RFC2671]  Vixie, P., "Extension Mechanisms for DNS (EDNS0)",
              RFC 2671, DOI 10.17487/RFC2671, August 1999,
              <https://www.rfc-editor.org/info/rfc2671>.

   [RFC2672]  Crawford, M., "Non-Terminal DNS Name Redirection",
              RFC 2672, DOI 10.17487/RFC2672, August 1999,
              <https://www.rfc-editor.org/info/rfc2672>.

   [RFC2673]  Crawford, M., "Binary Labels in the Domain Name System",
              RFC 2673, DOI 10.17487/RFC2673, August 1999,
              <https://www.rfc-editor.org/info/rfc2673>.

   [RFC3490]  Faltstrom, P., Hoffman, P., and A. Costello,
              "Internationalizing Domain Names in Applications (IDNA)",
              RFC 3490, DOI 10.17487/RFC3490, March 2003,
              <https://www.rfc-editor.org/info/rfc3490>.





Klensin                       Informational                    [Page 25]

RFC 8324                      DNS Revisions                February 2018


   [RFC3491]  Hoffman, P. and M. Blanchet, "Nameprep: A Stringprep
              Profile for Internationalized Domain Names (IDN)",
              RFC 3491, DOI 10.17487/RFC3491, March 2003,
              <https://www.rfc-editor.org/info/rfc3491>.

   [RFC3596]  Thomson, S., Huitema, C., Ksinant, V., and M. Souissi,
              "DNS Extensions to Support IP Version 6", STD 88,
              RFC 3596, DOI 10.17487/RFC3596, October 2003,
              <https://www.rfc-editor.org/info/rfc3596>.

   [RFC3743]  Konishi, K., Huang, K., Qian, H., and Y. Ko, "Joint
              Engineering Team (JET) Guidelines for Internationalized
              Domain Names (IDN) Registration and Administration for
              Chinese, Japanese, and Korean", RFC 3743,
              DOI 10.17487/RFC3743, April 2004,
              <https://www.rfc-editor.org/info/rfc3743>.

   [RFC3986]  Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform
              Resource Identifier (URI): Generic Syntax", STD 66,
              RFC 3986, DOI 10.17487/RFC3986, January 2005,
              <https://www.rfc-editor.org/info/rfc3986>.

   [RFC4033]  Arends, R., Austein, R., Larson, M., Massey, D., and S.
              Rose, "DNS Security Introduction and Requirements",
              RFC 4033, DOI 10.17487/RFC4033, March 2005,
              <https://www.rfc-editor.org/info/rfc4033>.

   [RFC4034]  Arends, R., Austein, R., Larson, M., Massey, D., and S.
              Rose, "Resource Records for the DNS Security Extensions",
              RFC 4034, DOI 10.17487/RFC4034, March 2005,
              <https://www.rfc-editor.org/info/rfc4034>.

   [RFC4343]  Eastlake 3rd, D., "Domain Name System (DNS) Case
              Insensitivity Clarification", RFC 4343,
              DOI 10.17487/RFC4343, January 2006,
              <https://www.rfc-editor.org/info/rfc4343>.

   [RFC5890]  Klensin, J., "Internationalized Domain Names for
              Applications (IDNA): Definitions and Document Framework",
              RFC 5890, DOI 10.17487/RFC5890, August 2010,
              <https://www.rfc-editor.org/info/rfc5890>.

   [RFC5891]  Klensin, J., "Internationalized Domain Names in
              Applications (IDNA): Protocol", RFC 5891,
              DOI 10.17487/RFC5891, August 2010,
              <https://www.rfc-editor.org/info/rfc5891>.





Klensin                       Informational                    [Page 26]

RFC 8324                      DNS Revisions                February 2018


   [RFC6672]  Rose, S. and W. Wijngaards, "DNAME Redirection in the
              DNS", RFC 6672, DOI 10.17487/RFC6672, June 2012,
              <https://www.rfc-editor.org/info/rfc6672>.

   [RFC6761]  Cheshire, S. and M. Krochmal, "Special-Use Domain Names",
              RFC 6761, DOI 10.17487/RFC6761, February 2013,
              <https://www.rfc-editor.org/info/rfc6761>.

   [RFC6891]  Damas, J., Graff, M., and P. Vixie, "Extension Mechanisms
              for DNS (EDNS(0))", STD 75, RFC 6891,
              DOI 10.17487/RFC6891, April 2013,
              <https://www.rfc-editor.org/info/rfc6891>.

   [RFC6895]  Eastlake 3rd, D., "Domain Name System (DNS) IANA
              Considerations", BCP 42, RFC 6895, DOI 10.17487/RFC6895,
              April 2013, <https://www.rfc-editor.org/info/rfc6895>.

   [RFC6912]  Sullivan, A., Thaler, D., Klensin, J., and O. Kolkman,
              "Principles for Unicode Code Point Inclusion in Labels in
              the DNS", RFC 6912, DOI 10.17487/RFC6912, April 2013,
              <https://www.rfc-editor.org/info/rfc6912>.

   [RFC7094]  McPherson, D., Oran, D., Thaler, D., and E. Osterweil,
              "Architectural Considerations of IP Anycast", RFC 7094,
              DOI 10.17487/RFC7094, January 2014,
              <https://www.rfc-editor.org/info/rfc7094>.

   [RFC7706]  Kumari, W. and P. Hoffman, "Decreasing Access Time to Root
              Servers by Running One on Loopback", RFC 7706,
              DOI 10.17487/RFC7706, November 2015,
              <https://www.rfc-editor.org/info/rfc7706>.

   [RFC7719]  Hoffman, P., Sullivan, A., and K. Fujiwara, "DNS
              Terminology", RFC 7719, DOI 10.17487/RFC7719, December
              2015, <https://www.rfc-editor.org/info/rfc7719>.

   [RFC7816]  Bortzmeyer, S., "DNS Query Name Minimisation to Improve
              Privacy", RFC 7816, DOI 10.17487/RFC7816, March 2016,
              <https://www.rfc-editor.org/info/rfc7816>.

   [RFC7858]  Hu, Z., Zhu, L., Heidemann, J., Mankin, A., Wessels, D.,
              and P. Hoffman, "Specification for DNS over Transport
              Layer Security (TLS)", RFC 7858, DOI 10.17487/RFC7858, May
              2016, <https://www.rfc-editor.org/info/rfc7858>.







Klensin                       Informational                    [Page 27]

RFC 8324                      DNS Revisions                February 2018


   [RFC7872]  Gont, F., Linkova, J., Chown, T., and W. Liu,
              "Observations on the Dropping of Packets with IPv6
              Extension Headers in the Real World", RFC 7872,
              DOI 10.17487/RFC7872, June 2016,
              <https://www.rfc-editor.org/info/rfc7872>.

   [RFC8244]  Lemon, T., Droms, R., and W. Kumari, "Special-Use Domain
              Names Problem Statement", RFC 8244, DOI 10.17487/RFC8244,
              October 2017, <https://www.rfc-editor.org/info/rfc8244>.

   [Sullivan-Class]
              Sullivan, A., "The DNS Is Not Classy: DNS Classes
              Considered Useless", Work in Progress,
              draft-sullivan-dns-class-useless-03, July 2016.

   [Unicode]  The Unicode Consortium, The Unicode Standard, Version
              9.0.0, (Mountain View, CA: The Unicode Consortium,
              2016. ISBN 978-1-936213-13-9),
              <http://www.unicode.org/versions/Unicode9.0.0/>.

   [Unicode-UAX15]
              Davis, M. and K. Whistler, "Unicode Standard Annex #15:
              Unicode Normalization Forms", February 2016,
              <http://unicode.org/reports/tr15/>.

   [Unicode-USA31]
              Davis, M., "Unicode Standard Annex #31: Unicode Identifier
              and Pattern Syntax", May 2016,
              <http://unicode.org/reports/tr31/>.

   [Vixie-20170704]
              Vixie, P., "Subject: Re: new DNS classes", message to
              the IETF dnsop mailing list, 4 July 2017,
              <https://www.ietf.org/mail-archive/web/ietf/current/
              msg103486.html>.
















Klensin                       Informational                    [Page 28]

RFC 8324                      DNS Revisions                February 2018


Acknowledgements



   Many of the concerns and ideas described in this document reflect
   conversations over a period of many years, some rooted in DNS
   "keyword" and "search" discussions that paralleled the development of
   IDNs.  Conversations with, or writings of, Rob Austein, Christine
   Borgman, Carolina Carvalho, Vint Cerf, Lyman Chapin, Nazli Choucri,
   Patrik Faltstrom, Geoff Huston, Xiaodong Lee, Karen Liu, Gervase
   Markham, Yaqub Mueller, Andrew Sullivan, Paul Twomey, Nico Williams,
   Suzanne Woolf, Jiankang Yao, other participants in the circa 2003
   "DNS Search" effort and in the ICANN SSAC Working Party on IDNs, and
   some others whose names were sadly forgotten, were particularly
   important to either the content of this document or the motivation
   for writing it even though they may not agree with the conclusions I
   have reached and bear no responsibility for them.

   Many of the subsections of Section 3 were extracted from comments
   first made in conjunction with recent email discussions.  Comments
   from Suzanne Woolf about an earlier draft version were particularly
   important as was material developed with suggestions from Patrik
   Faltstrom, especially Section 3.13.  Feedback and suggestions from
   several of the above and from Stephane Bortzmeyer, Tony Finch, Bob
   Harold, Warren Kumari, Craig Partridge, and George Sadowsky were
   extremely helpful for improving the clarity and accuracy of parts of
   the document, especially so for a broader audience.  Craig Partridge
   also contributed much of the material about queries for multiple
   types.  Geoff Huston made several useful comments and contributed
   most of Section 3.15, and Bill Manning pointed out some broader
   requirements about integrity of information and DNS management and
   operations.

   Special thanks are due to Karen Moore of the RFC Production Center
   for her efforts, patience, and persistence in preparing this document
   for publication, a process that raised far more issues that required
   careful discussion than usual.

Author's Address



   John C. Klensin
   1770 Massachusetts Ave, Ste 322
   Cambridge, MA  02140
   United States of America

   Phone: +1 617 245 1457
   Email: john-ietf@jck.com






Klensin                       Informational                    [Page 29]