RFC 7108






Independent Submission                                          J. Abley
Request for Comments: 7108                                     Dyn, Inc.
Category: Informational                                     T. Manderson
ISSN: 2070-1721                                                    ICANN
                                                            January 2014


       A Summary of Various Mechanisms Deployed at L-Root for the
                    Identification of Anycast Nodes

Abstract



   Anycast is a deployment technique commonly employed for
   authoritative-only servers in the Domain Name System (DNS).  L-Root,
   one of the thirteen root servers, is deployed in this fashion.

   Various techniques have been used to map deployed anycast
   infrastructure externally, i.e., without reference to inside
   knowledge about where and how such infrastructure has been deployed.
   Motivations for performing such measurement exercises include
   operational troubleshooting and infrastructure risk assessment.  In
   the specific case of L-Root, the ability to measure and map anycast
   infrastructure using the techniques mentioned in this document is
   provided for reasons of operational transparency.

   This document describes all facilities deployed at L-Root to
   facilitate mapping of its infrastructure and serves as documentation
   for L-Root as a measurable service.

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 a candidate for any level of Internet
   Standard; see Section 2 of RFC 5741.

   Information about the current status of this document, any errata,
   and how to provide feedback on it may be obtained at
   http://www.rfc-editor.org/info/rfc7108.







Abley & Manderson             Informational                     [Page 1]

RFC 7108           L-Root Anycast Node Identification       January 2014


Copyright Notice



   Copyright (c) 2014 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
   (http://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.

Table of Contents



   1.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   2
   2.  Conventions Used in This Document . . . . . . . . . . . . . .   3
   3.  Naming Scheme for L-Root Nodes  . . . . . . . . . . . . . . .   3
   4.  Identification of L-Root Nodes  . . . . . . . . . . . . . . .   3
     4.1.  Use of NSID . . . . . . . . . . . . . . . . . . . . . . .   4
     4.2.  Use of HOSTNAME.BIND/CH/TXT . . . . . . . . . . . . . . .   5
     4.3.  Use of ID.SERVER/CH/TXT . . . . . . . . . . . . . . . . .   6
     4.4.  Use of IDENTITY.L.ROOT-SERVERS.ORG/IN/TXT and .../IN/A  .   6
     4.5.  Use of NODES.L.ROOT-SERVERS.ORG/IN/TXT  . . . . . . . . .   8
   5.  Provisioning of IDENTITY.L.ROOT-SERVERS.ORG . . . . . . . . .   9
   6.  Security Considerations . . . . . . . . . . . . . . . . . . .   9
   7.  Acknowledgements  . . . . . . . . . . . . . . . . . . . . . .  10
   8.  References  . . . . . . . . . . . . . . . . . . . . . . . . .  10
     8.1.  Normative References  . . . . . . . . . . . . . . . . . .  10
     8.2.  Informative References  . . . . . . . . . . . . . . . . .  11

1.  Introduction



   The Domain Name System (DNS) is described in [RFC1034] and [RFC1035].
   L-Root, one of the thirteen root servers, is deployed using anycast
   [RFC4786]; its service addresses, published in the A and AAAA
   Resource Record (RR) Sets for "L.ROOT-SERVERS.NET", are made
   available from a substantial number of semi-autonomous servers
   deployed throughout the Internet.  A list of locations served by
   L-Root can be found at [ROOT-SERVERS].

   Various techniques have been used to map deployed anycast
   infrastructure externally, i.e., without reference to inside
   knowledge about where and how such infrastructure has been deployed.
   Motivations for performing such measurement exercises include
   operational troubleshooting and infrastructure risk assessment.  In
   the specific case of L-Root, the ability to measure and map anycast
   infrastructure using the techniques mentioned in this document is
   provided for reasons of operational transparency.



Abley & Manderson             Informational                     [Page 2]

RFC 7108           L-Root Anycast Node Identification       January 2014


   This document describes all facilities currently provided at L-Root
   to aid node identification.

2.  Conventions Used in This Document



   This document contains several examples of commands typed at a Unix
   (or Unix-like) command line to illustrate use of the various
   mechanisms available to identify L-Root nodes.  Such examples are
   presented in this document with lines typed by the user preceded by
   the "%" prompt character; a bare "%" character indicates the end of
   the output resulting from the command.

   In some cases, the output shown in examples is too long to be
   represented directly in the text.  In those cases, a backslash
   character ("\") is used to indicate continuation.

3.  Naming Scheme for L-Root Nodes



   Individual L-Root nodes have structured hostnames that are
   constructed as follows:

      <IATA Code><NN>.L.ROOT-SERVERS.ORG

   where

   o  <IATA Code> is chosen from the list of three-character airport
      codes published by the International Air Transport Association
      (IATA) in the IATA Airline Coding Directory [ACD]; and

   o  <NN> is a two-digit numeric code used to distinguish between two
      different nodes in the vicinity of the same airport.

   Where multiple airports exist in the vicinity of a single L-Root
   node, one is arbitrarily chosen.

   More granular location data published for L-Root nodes (e.g., see
   Section 4.4) is derived from the location of the airport, not the
   actual location of the node.

4.  Identification of L-Root Nodes



   L-Root service is provided using a single IPv4 address (199.7.83.42)
   and a single IPv6 address (2001:500:3::42).  Note that it is
   preferable to refer to the service using its DNS name (L.ROOT-
   SERVERS.NET) rather than literal addresses, since addresses can
   change from time to time.





Abley & Manderson             Informational                     [Page 3]

RFC 7108           L-Root Anycast Node Identification       January 2014


   At the time of writing, there are 273 separate name server elements
   ("nodes") deployed in 143 locations: together, these nodes provide
   L-Root service.  A DNS query sent to an L-Root service address will
   be routed towards exactly one of those nodes for processing, and the
   corresponding DNS response will be originated from the same node.
   Queries from different clients may be routed to different nodes.
   Successive queries from the same client may also be routed to
   different nodes.

   The following sections provide a summary of all mechanisms provided
   by L-Root to allow a client to identify which L-Root node is being
   used.

   Using HOSTNAME.BIND/CH/TXT (Section 4.2), ID.SERVER/CH/TXT
   (Section 4.3), or IDENTITY.L.ROOT-SERVERS.ORG/IN/TXT or IDENTITY.L
   .ROOT-SERVERS/IN/A (Section 4.4) to identify a node for the purposes
   of reporting a problem is frequently reasonable, but it should be
   acknowledged that there is potential for re-routing between
   successive queries: an observed problem might relate to one node,
   whilst a subsequent query using one of those three techniques could
   be answered by a different node.  Use of the Name Server Identifier
   (NSID) option on the precise queries that yield problematic responses
   can obviate this possibility (see Section 4.1).

4.1.  Use of NSID



   L-Root supports the use of the Name Server Identifier (NSID) option
   [RFC5001] to return the identity of an L-Root node along with the
   response to a DNS query.  The NSID payload of such responses is the
   fully qualified hostname of the responding L-Root node.

   The NSID option allows the identification of a node sending a
   specific, requested response to the client.  This is of particular
   use if (for example) there is a desire to identify unequivocally what
   node is responding with a particularly troublesome response; the
   output of the diagnostic tool "dig" with NSID requested provides the
   problem response with the node identification, and its output in that
   case could form the basis of a useful trouble report.

   NSID is specified as an EDNS(0) option [RFC6891].  Clients that do
   not support EDNS(0) signaling (or depend on other systems that do not
   support EDNS0) may find this mechanism unavailable.

   The NSID option can be specified using the widely used diagnostic
   tool "dig" using the "+nsid" option, as shown below.  Note that long
   lines have been truncated for the purposes of this document ("\" at
   the end of a line indicates continuation).




Abley & Manderson             Informational                     [Page 4]

RFC 7108           L-Root Anycast Node Identification       January 2014


   % dig -4 @L.ROOT-SERVERS.NET . SOA +nsid \
     +norec +noall +comments
   ; <<>> DiG 9.6.-ESV-R3 <<>> -4 @L.ROOT-SERVERS.NET . SOA +nsid \
     +norec +noall +comments
   ; (1 server found)
   ;; global options: +cmd
   ;; Got answer:
   ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 14913
   ;; flags: qr aa; QUERY: 1, ANSWER: 1, AUTHORITY: 13, ADDITIONAL: 23

   ;; OPT PSEUDOSECTION:
   ; EDNS: version: 0, flags:; udp: 4096
   ; NSID: 79 74 7a 30 31 2e 6c 2e 72 6f 6f 74 2d 73 65 72 76 65 72 73 \
     2e 6f 72 67  (y) (t) (z) (0) (1) (.) (l) (.) (r) (o) (o) (t) (-) \
     (s) (e) (r) (v) (e) (r) (s) (.) (o) (r) (g)
   %

   % dig -6 @L.ROOT-SERVERS.NET . SOA +nsid \
     +norec +noall +comments
   ; <<>> DiG 9.6.-ESV-R3 <<>> -6 @L.ROOT-SERVERS.NET . SOA +nsid \
     +norec +noall +comments
   ; (1 server found)
   ;; global options: +cmd
   ;; Got answer:
   ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 33374
   ;; flags: qr aa; QUERY: 1, ANSWER: 1, AUTHORITY: 13, ADDITIONAL: 23

   ;; OPT PSEUDOSECTION:
   ; EDNS: version: 0, flags:; udp: 4096
   ; NSID: 79 74 7a 30 31 2e 6c 2e 72 6f 6f 74 2d 73 65 72 76 65 72 73 \
     2e 6f 72 67  (y) (t) (z) (0) (1) (.) (l) (.) (r) (o) (o) (t) (-) \
     (s) (e) (r) (v) (e) (r) (s) (.) (o) (r) (g)
   %

4.2.  Use of HOSTNAME.BIND/CH/TXT



   L-Root supports the use of HOSTNAME.BIND/CH/TXT queries to return the
   identity of an L-Root node.  The TXT RDATA returned is the fully
   qualified hostname of the responding L-Root node.

   The HOSTNAME.BIND/CH/TXT convention is described in [RFC4892].










Abley & Manderson             Informational                     [Page 5]

RFC 7108           L-Root Anycast Node Identification       January 2014


   % dig -4 @L.ROOT-SERVERS.NET HOSTNAME.BIND CH TXT +short
   "ytz01.l.root-servers.org"
   %

   % dig -6 @L.ROOT-SERVERS.NET HOSTNAME.BIND CH TXT +short
   "ytz01.l.root-servers.org"
   %

4.3.  Use of ID.SERVER/CH/TXT



   L-Root supports the use of ID.SERVER/CH/TXT queries to return the
   identity of an L-Root node.  The TXT RDATA returned is the fully
   qualified hostname of the responding L-Root node.

   ID.SERVER/CH/TXT functions identically (apart from the QNAME) to
   HOSTNAME.BIND/CH/TXT, as discussed in Section 4.2.  The discussion
   there relating to the possibility of re-routing between successive
   queries also follows for ID.SERVER/CH/TXT.

   The ID.SERVER/CH/TXT convention is described in [RFC4892].

   % dig -4 @L.ROOT-SERVERS.NET ID.SERVER CH TXT +short
   "ytz01.l.root-servers.org"
   %

   % dig -6 @L.ROOT-SERVERS.NET ID.SERVER CH TXT +short
   "ytz01.l.root-servers.org"
   %

4.4.  Use of IDENTITY.L.ROOT-SERVERS.ORG/IN/TXT and .../IN/A

   The operator of L-Root has distributed a separate DNS service in
   parallel with L-Root, operating on precisely the same set of nodes
   but listening on addresses that are different from the L-Root service
   addresses.  Measurements of this separate service should give results
   that are representative of L-Root.  Further discussion of this
   service can be found in Section 5.

   The fully qualified DNS name IDENTITY.L.ROOT-SERVERS.ORG (note the
   use of ORG, not NET) has associated TXT and A RR Sets that are unique
   to the responding node.  Clients are hence able to issue queries for
   IDENTITY.L.ROOT-SERVERS.ORG/IN/A and IDENTITY.L.ROOT-SERVERS.ORG/IN/
   TXT and use the results both to identify individual nodes and to
   distinguish between responses generated by different nodes.







Abley & Manderson             Informational                     [Page 6]

RFC 7108           L-Root Anycast Node Identification       January 2014


   The TXT record returned in the response to such queries is structured
   as follows:

   1.  The fully qualified hostname of the node responding to the query;

   2.  The city in which the node is located;

   3.  The region in which the node is located, if applicable;

   4.  The economy in which the node is located (in most cases, the name
       of a country); and

   5.  The Internet Corporation for Assigned Names and Numbers (ICANN)
       region in which the node is located.  A list of ICANN regions at
       the time of writing can be found at <http://meetings.icann.org/
       regions>.

   The A record returned in the response to such queries is guaranteed
   to be unique to the responding node.  The A RRType was chosen in an
   effort to make the use of this mechanism as widely available to
   client environments as possible, and the ability to map a hostname to
   an IPv4 address seemed more likely to be widespread than the mapping
   of a hostname to any other value.  It should be noted that the
   availability of this mechanism to any particular client is orthogonal
   to the local availability of IPv4 or IPv6 transport.

   In this case, because identity data is published using IN-class
   resource records, it is not necessary to send queries directly
   towards L-Root in order to obtain results.  Responses can be obtained
   through recursive servers, the responses in those cases being the
   identity of L-Root as observed through the recursive server used
   rather than the "closest" L-Root node to the client.  This
   facilitates some degree of remote troubleshooting, since a query for
   IDENTITY.L.ROOT-SERVERS.ORG/IN/TXT or IDENTITY.L.ROOT-SERVERS.ORG/IN/
   A directed a remote recursive resolver can help illustrate which
   L-Root node is being used by that server (or was used when the cache
   was populated).

   A related caching effect is that responses to IDENTITY.L.ROOT-
   SERVERS.ORG/IN/A and IDENTITY.L.ROOT-SERVERS.ORG/IN/TXT may be cached
   at different times, and may hence persist in a cache for overlapping
   periods of time.  One possible visible effect is that the responses
   to IDENTITY.L.ROOT-SERVERS.ORG/IN/A and IDENTITY.L.ROOT-SERVERS.ORG/
   IN/TXT as presented from a cache may appear to be incoherent (i.e.,
   refer to different nodes) despite queries against of the cache
   happening (near) simultaneously.  Caches may also discard the
   published Times to Live (TTLs) in responses from the authoritative




Abley & Manderson             Informational                     [Page 7]

RFC 7108           L-Root Anycast Node Identification       January 2014


   server and replace them with longer TTLs, as a matter of local
   policy.  Interpretation of responses for these queries from caches
   should therefore be carried out with these possible effects in mind.

   It has been observed that IDENTITY.L.ROOT-SERVERS.ORG/IN/A queries
   offer a useful mechanism for troubleshooting DNS problems with non-
   technical users, since such users can often be walked through the
   process of looking up an A record (e.g., as a side effect of
   utilities such as ping) far easier than they can be instructed on how
   to use DNS-specific tools such as dig.

  % dig IDENTITY.L.ROOT-SERVERS.ORG TXT +short
  "ytz01.l.root-servers.org" "Toronto" "Ontario" "Canada" "NorthAmerica"
  %

  % dig IDENTITY.L.ROOT-SERVERS.ORG A +short
  67.215.199.91
  %

4.5.  Use of NODES.L.ROOT-SERVERS.ORG/IN/TXT



   The fully qualified DNS name NODES.L.ROOT-SERVERS.ORG (note again the
   use of ORG, not NET) provides multiple TXT RRs, one per node, and
   represents the effective concatenation of all possible responses to
   the query IDENTITY.L.ROOT-SERVERS.ORG/IN/TXT.

   Note that in the example below we have forced dig to send the query
   over TCP, since we expect the response to be too large for UDP
   transport to accommodate.  Note also that the list shown is truncated
   for clarity, and can be expected to change from time to time as new
   L-Root nodes are provisioned and old ones decommissioned.

   % dig NODES.L.ROOT-SERVERS.ORG TXT +short +tcp | head -10
   "abj01.l.root-servers.org" "Abidjan" "" "Cote d'Ivoire" "Africa"
   "abj02.l.root-servers.org" "Abidjan" "" "Cote d'Ivoire" "Africa"
   "akl01.l.root-servers.org" "Mangere" "" "New Zealand" "AsiaPacific"
   "akl41.l.root-servers.org" "Mangere" "" "New Zealand" "AsiaPacific"
   "akl42.l.root-servers.org" "Mangere" "" "New Zealand" "AsiaPacific"
   "akl43.l.root-servers.org" "Mangere" "" "New Zealand" "AsiaPacific"
   "akl44.l.root-servers.org" "Mangere" "" "New Zealand" "AsiaPacific"
   "ams01.l.root-servers.org" "Haarlemmermeer" "" "Netherlands" "Europe"
   "anc01.l.root-servers.org" "Anchorage" "Alaska" "United States" \
     "NorthAmerica"
   %







Abley & Manderson             Informational                     [Page 8]

RFC 7108           L-Root Anycast Node Identification       January 2014


5.  Provisioning of IDENTITY.L.ROOT-SERVERS.ORG



   Individual L-Root nodes run a dedicated, separate authority-only DNS
   server process that serves the IDENTITY.L.ROOT-SERVERS.ORG zone.  The
   contents of that zone are unique to every node; hence, each
   responding node will generate a node-specific response.

   The contents of the IDENTITY.L.ROOT-SERVERS.ORG zone are hence
   deliberately incoherent, the apparent zone contents depending on the
   node responding to the corresponding query.

   The IDENTITY.L.ROOT-SERVERS.ORG zone is delegated to the single name
   server BEACON.L.ROOT-SERVERS.ORG, numbered on IPv4 and IPv6 addresses
   that are covered by the same routing advertisements that cover the
   L-Root service addresses.  Reachability of BEACON.L.ROOT-SERVERS.ORG
   is hence well-aligned with the reachability of L.ROOT-SERVERS.NET;
   therefore, measurement of the IDENTITY service ought to give similar
   results to measurement of the L-Root service.

   It is considered best practice always to delegate a DNS zone to more
   than one name server [RFC2182]; however, as described, the IDENTITY.L
   .ROOT-SERVERS.ORG zone is delegated to just one server.  Ordinarily,
   this would present a risk of failure if that single server is not
   available; however, given the purpose of the delegation in this case
   and that the expected mitigation of a failure in a single node is the
   routing of a query to a different node, delegation to a single server
   in this particular use-case is effective.

   At the time of writing, the ROOT-SERVERS.ORG zone is not signed with
   DNSSEC.  When DNSSEC is deployed in that zone, the L.ROOT-SERVERS.ORG
   zone will also be signed.  This will facilitate secure responses for
   queries for BEACON.L.ROOT-SERVERS.ORG and NODES.L.ROOT-SERVERS.ORG.

   Secure responses for IDENTITY.L.ROOT-SERVERS.ORG are unlikely to
   become available even with the deployment of DNSSEC in the parent,
   since the implementation of the IDENTITY.L.ROOT-SERVERS.ORG service
   involves widely distributed static zone data.  Management of key
   materials distributed to every L-Root node would be impractical to
   audit, and signatures returned in secure responses would be
   correspondingly of low value.

6.  Security Considerations



   Some operators of anycast services choose not to disclose locations
   (or even numbers) of nodes, citing security concerns.  The operator
   of L-Root considers that none of the published information described
   in this document is truly secret, since any service element that
   provides service to the Internet can never truly be obscured from



Abley & Manderson             Informational                     [Page 9]

RFC 7108           L-Root Anycast Node Identification       January 2014


   view.  Given that location information can be found regardless of any
   conscious, deliberate disclosure, and since easy access to this
   information has diagnostic value, the operator of L-Root has adopted
   a policy of operational transparency.

   The information presented in this document presents no new threat to
   the Internet.

7.  Acknowledgements



   The aspects of the L-Root service that were deployed to facilitate
   IN-class mapping were discussed and implemented as part of an
   informal collaboration with Xun Fan, John Heidemann, and Ramesh
   Govidan, whose contributions are acknowledged.  The motivation to
   facilitate mapping of L-Root as an anycast service using IN-class
   queries was inspired by [Fan2013].

   Helpful reviews and comments from Gaurab Upadhaya, Hugo Salgado,
   Brian Dixon, Bob Harold, Paul Hoffman, Jakob Schlyter, Andrew
   Sullivan, Bruce Campbell, S. Moonesamy, and Stephane Bortzmeyer on
   earlier versions of this document were very much appreciated.

8.  References



8.1.  Normative References



   [RFC1034]  Mockapetris, P., "Domain names - concepts and facilities",
              STD 13, RFC 1034, November 1987.

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

   [RFC2182]  Elz, R., Bush, R., Bradner, S., and M. Patton, "Selection
              and Operation of Secondary DNS Servers", BCP 16, RFC 2182,
              July 1997.

   [RFC4786]  Abley, J. and K. Lindqvist, "Operation of Anycast
              Services", BCP 126, RFC 4786, December 2006.

   [RFC4892]  Woolf, S. and D. Conrad, "Requirements for a Mechanism
              Identifying a Name Server Instance", RFC 4892, June 2007.

   [RFC5001]  Austein, R., "DNS Name Server Identifier (NSID) Option",
              RFC 5001, August 2007.

   [RFC6891]  Damas, J., Graff, M., and P. Vixie, "Extension Mechanisms
              for DNS (EDNS(0))", STD 75, RFC 6891, April 2013.




Abley & Manderson             Informational                    [Page 10]

RFC 7108           L-Root Anycast Node Identification       January 2014


8.2.  Informative References



   [ACD]      International Air Transport Association (IATA), "Airline
              Coding Directory (ACD)", 2013,
              <http://www.iata.org/publications/Pages/coding.aspx>.

   [Fan2013]  Fan, X., Heidemann, J., and R. Govidan, "Evaluating
              Anycast in the Domain Name System", Proceedings of the
              IEEE Infocom Turin, Italy, April 2013.

   [ROOT-SERVERS]
              "root-servers.org", <http://www.root-servers.org>.

Authors' Addresses



   Joe Abley
   Dyn, Inc.
   470 Moore Street
   London, ON  N6C 2C2
   Canada

   Phone: +1 519 670 9327
   EMail: jabley@dyn.com


   Terry Manderson
   ICANN
   12025 Waterfront Drive
   Suite 300
   Los Angeles, CA  90094-2536
   USA

   EMail: terry.manderson@icann.org


















Abley & Manderson             Informational                    [Page 11]