RFC 8777




Internet Engineering Task Force (IETF)                        J. Holland
Request for Comments: 8777                     Akamai Technologies, Inc.
Updates: 7450                                                 April 2020
Category: Standards Track
ISSN: 2070-1721


      DNS Reverse IP Automatic Multicast Tunneling (AMT) Discovery

Abstract



   This document updates RFC 7450, "Automatic Multicast Tunneling" (or
   AMT), by modifying the relay discovery process.  A new DNS resource
   record named AMTRELAY is defined for publishing AMT relays for
   source-specific multicast channels.  The reverse IP DNS zone for a
   multicast sender's IP address is configured to use AMTRELAY resource
   records to advertise a set of AMT relays that can receive and forward
   multicast traffic from that sender over an AMT tunnel.  Other
   extensions and clarifications to the relay discovery process are also
   defined.

Status of This Memo



   This is an Internet Standards Track document.

   This document is a product of the Internet Engineering Task Force
   (IETF).  It represents the consensus of the IETF community.  It has
   received public review and has been approved for publication by the
   Internet Engineering Steering Group (IESG).  Further information on
   Internet Standards is available in 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/rfc8777.

Copyright Notice



   Copyright (c) 2020 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.  Code Components extracted from this document must
   include Simplified BSD License text as described in Section 4.e of
   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.

Table of Contents



   1.  Introduction
     1.1.  Background
     1.2.  Terminology
       1.2.1.  Relays and Gateways
       1.2.2.  Definitions
       1.2.3.  Requirements Language
   2.  Relay Discovery Overview
     2.1.  Basic Mechanics
     2.2.  Signaling and Discovery
     2.3.  Example Deployments
       2.3.1.  Example Receiving Networks
       2.3.2.  Example Sending Networks
   3.  Relay Discovery Operation
     3.1.  Optimal Relay Selection
       3.1.1.  Overview
       3.1.2.  Preference Ordering
       3.1.3.  Connecting to Multiple Relays
     3.2.  Happy Eyeballs
       3.2.1.  Overview
       3.2.2.  Algorithm Guidelines
       3.2.3.  Connection Definition
     3.3.  Guidelines for Restarting Discovery
       3.3.1.  Overview
       3.3.2.  Updates to Restarting Events
       3.3.3.  Tunnel Stability
       3.3.4.  Traffic Health
       3.3.5.  Relay Loaded or Shutting Down
       3.3.6.  Relay Discovery Messages vs. Restarting Discovery
       3.3.7.  Independent Discovery per Traffic Source
     3.4.  DNS Configuration
     3.5.  Waiting for DNS Resolution
   4.  AMTRELAY Resource Record Definition
     4.1.  AMTRELAY RRType
     4.2.  AMTRELAY RData Format
       4.2.1.  RData Format - Precedence
       4.2.2.  RData Format - Discovery Optional (D-bit)
       4.2.3.  RData Format - Type
       4.2.4.  RData Format - Relay
     4.3.  AMTRELAY Record Presentation Format
       4.3.1.  Representation of AMTRELAY RRs
       4.3.2.  Examples
   5.  IANA Considerations
   6.  Security Considerations
     6.1.  Use of AMT
     6.2.  Record-Spoofing
     6.3.  Congestion
   7.  References
     7.1.  Normative References
     7.2.  Informative References
   Appendix A.  Unknown RRType Construction
   Acknowledgements

   Author's Address



1.  Introduction



   This document defines DNS Reverse IP AMT Discovery (DRIAD), a
   mechanism for AMT gateways to discover AMT relays that are capable of
   forwarding multicast traffic from a known source IP address.

   AMT (Automatic Multicast Tunneling) is defined in [RFC7450] and
   provides a method to transport multicast traffic over a unicast
   tunnel in order to traverse network segments that are not multicast
   capable.

   Section 4.1.5 of [RFC7450] explains that the relay selection process
   for AMT is intended to be more flexible than the particular discovery
   method described in that document.  That section further explains
   that the selection process might need to depend on the source of the
   multicast traffic in some deployments, since a relay must be able to
   receive multicast traffic from the desired source in order to forward
   it.

   Section 4.1.5 of [RFC7450] goes on to suggest DNS-based queries as a
   possible solution: DRIAD is DNS based.  This solution also addresses
   the relay discovery issues in the "Disadvantages of this
   configuration" lists in Sections 3.3 and 3.4 of [RFC8313].

   The goal for DRIAD is to enable multicast connectivity between
   separate multicast-enabled networks without preconfiguring any
   peering arrangements between the networks when neither the sending
   nor the receiving network is connected to a multicast-enabled
   backbone.

   This document extends the relay discovery procedure described in
   Section 5.2.3.4 of [RFC7450].

1.1.  Background



   The reader is assumed to be familiar with the basic DNS concepts
   described in [RFC1034], [RFC1035], and the subsequent documents that
   update them, particularly [RFC2181].

   The reader is also assumed to be familiar with the concepts and
   terminology regarding source-specific multicast as described in
   [RFC4607] and the use of Internet Group Management Protocol Version 3
   (IGMPv3) [RFC3376] and Multicast Listener Discovery Version 2 (MLDv2)
   [RFC3810] for group management of source-specific multicast channels,
   as described in [RFC4604].

   The reader should also be familiar with AMT, particularly the
   terminology listed in Sections 3.2 and 3.3 of [RFC7450].

1.2.  Terminology



1.2.1.  Relays and Gateways



   When reading this document, it's especially helpful to recall that
   once an AMT tunnel is established, the relay receives native
   multicast traffic and sends unicast tunnel-encapsulated traffic to
   the gateway.  The gateway receives the tunnel-encapsulated packets,
   decapsulates them, and forwards them as native multicast packets, as
   illustrated in Figure 1.

     Multicast  +-----------+  Unicast  +-------------+  Multicast
    >---------> | AMT relay | >=======> | AMT gateway | >--------->
                +-----------+           +-------------+

                     Figure 1: AMT Tunnel Illustration

1.2.2.  Definitions



     +------------+-------------------------------------------------+
     |       Term | Definition                                      |
     +============+=================================================+
     |      (S,G) | A source-specific multicast channel, as         |
     |            | described in [RFC4607].  A pair of IP addresses |
     |            | with a source host IP and destination group IP. |
     +------------+-------------------------------------------------+
     |       CMTS | Cable Modem Termination System                  |
     +------------+-------------------------------------------------+
     |  discovery | A broker or load balancer for AMT relay         |
     |     broker | discovery, as mentioned in Section 4.2.1.1 of   |
     |            | [RFC7450].                                      |
     +------------+-------------------------------------------------+
     | downstream | Further from the source of traffic, as          |
     |            | described in [RFC7450].                         |
     +------------+-------------------------------------------------+
     |       FQDN | Fully Qualified Domain Name, as described in    |
     |            | [RFC8499].                                      |
     +------------+-------------------------------------------------+
     |    gateway | An AMT gateway, as described in [RFC7450].      |
     +------------+-------------------------------------------------+
     |     L flag | The "Limit" flag described in Section 5.1.4.4   |
     |            | of [RFC7450].                                   |
     +------------+-------------------------------------------------+
     |        OLT | Optical Line Terminal                           |
     +------------+-------------------------------------------------+
     |      relay | An AMT relay, as described in [RFC7450].        |
     +------------+-------------------------------------------------+
     |        RPF | Reverse Path Forwarding, as described in        |
     |            | [RFC5110].                                      |
     +------------+-------------------------------------------------+
     |         RR | A DNS Resource Record, as described in          |
     |            | [RFC1034].                                      |
     +------------+-------------------------------------------------+
     |     RRType | A DNS Resource Record Type, as described in     |
     |            | [RFC1034].                                      |
     +------------+-------------------------------------------------+
     |        SSM | Source-specific multicast, as described in      |
     |            | [RFC4607].                                      |
     +------------+-------------------------------------------------+
     |   upstream | Closer to the source of traffic, as described   |
     |            | in [RFC7450].                                   |
     +------------+-------------------------------------------------+

                           Table 1: Definitions

1.2.3.  Requirements Language



   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
   "OPTIONAL" in this document are to be interpreted as described in
   BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all
   capitals, as shown here.

2.  Relay Discovery Overview



2.1.  Basic Mechanics



   The AMTRELAY resource record (RR) defined in this document is used to
   publish the IP address or domain name of a set of AMT relays or
   discovery brokers that can receive, encapsulate, and forward
   multicast traffic from a particular sender.

   The sender is the owner of the RR and configures the zone so that it
   contains a set of RRs that provide the addresses or domain names of
   AMT relays (or discovery brokers that advertise relays) that can
   receive multicast IP traffic from that sender.

   This enables AMT gateways in remote networks to discover an AMT relay
   that is capable of forwarding traffic from the sender.  This, in
   turn, enables those AMT gateways to receive the multicast traffic
   tunneled over a unicast AMT tunnel from those relays and then pass
   the multicast packets into networks or applications that are using
   the gateway to subscribe to traffic from that sender.

   This mechanism only works for source-specific multicast (SSM)
   channels.  The source address of the (S,G) is reversed and used as an
   index into one of the reverse mapping trees (in-addr.arpa for IPv4,
   as described in Section 3.5 of [RFC1035], or ip6.arpa for IPv6, as
   described in Section 2.5 of [RFC3596]).

   This mechanism should be treated as an extension of the AMT relay
   discovery procedure described in Section 5.2.3.4 of [RFC7450].  A
   gateway that supports this method of AMT relay discovery SHOULD use
   this method whenever it's performing the relay discovery procedure,
   the source IP addresses for desired (S,G)s are known to the gateway,
   and conditions match the requirements outlined in Section 3.1.

   Some detailed example use cases are provided in Section 2.3, and
   other applicable example topologies appear in Sections 3.3, 3.4, and
   3.5 of [RFC8313].

2.2.  Signaling and Discovery



   This section describes a typical example of the end-to-end process
   for signaling a receiver's join of an SSM channel that relies on an
   AMTRELAY RR.

   The example in Figure 2 contains two multicast-enabled networks that
   are both connected to the internet with non-multicast-capable links
   and which have no direct association with each other.

   A content provider operates a sender, which is a source of multicast
   traffic inside a multicast-capable network.

   An end user who is a customer of the content provider has a
   multicast-capable Internet Service Provider (ISP), which operates a
   receiving network that uses an AMT gateway.  The AMT gateway is DRIAD
   capable.

   The content provider provides the user with a receiving application
   that tries to subscribe to at least one (S,G).  This receiving
   application could, for example, be a file transfer system using File
   Delivery over Unidirectional Transport (FLUTE) [RFC6726], a live
   video stream using RTP [RFC3550], or any other application that might
   subscribe to an SSM channel.

                     +---------------+
                     |    Sender     |
      |    |         |  2001:db8::a  |
      |    |         +---------------+
      |Data|                 |
      |Flow|      Multicast  |
     \|    |/      Network   |
      \    /                 |        5: Propagate RPF for Join(S,G)
       \  /          +---------------+
        \/           |   AMT relay   |
                     | 2001:db8:c::f |
                     +---------------+
                             |        4: Gateway connects to Relay,
                                         sends Join(S,G) over tunnel
                             |
                    Unicast
                     Tunnel  |        3: --> DNS Query: type=AMTRELAY,
                                     /        a.0.0.0.0.0.0.0.0.0.0.0.
         ^                   |      /         0.0.0.0.0.0.0.0.0.0.0.0.
         |                         /          8.b.d.0.1.0.0.2.ip6.arpa
         |                   |    /      <-- Response:
     Join/Leave       +-------------+         AMTRELAY=2001:db8:c::f
      Signals         | AMT gateway |
         |            +-------------+
         |                   |        2: Propagate RPF for Join(S,G)
         |        Multicast  |
                   Network   |
                             |     1: Join(S=2001:db8::a,G=ff3e::8000:d)
                      +-------------+
                      |   Receiver  |
                      |  (end user) |
                      +-------------+

                         Figure 2: DRIAD Messaging

   In this simple example, the sender IP is 2001:db8::a, which is
   sending traffic to the group address ff3e::8000:d, and the relay IP
   is 2001:db8::c:f.

   The content provider has previously configured the DNS zone that
   contains the reverse IP domain name for the sender's IP address so
   that it provides an AMTRELAY RR with the relay's IP address (see
   Section 4.3 for details about the AMTRELAY RR format and semantics).
   As described in Section 2.5 of [RFC3596], the reverse IP FQDN of the
   sender's address "2001:db8::a" is:

    a.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.8.b.d.0.1.0.0.2.ip6.
                                                                   arpa.

   The sequence of events depicted in Figure 2 is as follows:

   1.  The end user starts the app, which issues a join to the (S,G):
       (2001:db8::a, ff3e::8000:d).

   2.  The join propagates with RPF through the receiver's multicast-
       enabled network with PIM [RFC7761] or another multicast routing
       mechanism until the AMT gateway receives a signal to join the
       (S,G).

   3.  The AMT gateway performs a reverse DNS lookup for the AMTRELAY
       RRType by sending an AMTRELAY RRType query for the reverse IP
       domain name for the sender's source IP address (the S from the
       (S,G)).

       The DNS resolver for the AMT gateway uses ordinary DNS recursive
       resolution until it has the authoritative result that the content
       provider configured, which informs the AMT gateway that the relay
       address is 2001:db8::c:f.

   4.  The AMT gateway performs AMT handshakes with the AMT relay as
       described in Section 4 of [RFC7450], then forwards a membership
       report to the relay, indicating subscription to the (S,G).

   5.  The relay propagates the join through its network toward the
       sender and then forwards the appropriate AMT-encapsulated traffic
       to the gateway, which decapsulates and forwards it as a native
       multicast through its downstream network to the end user.

   In the case of an IPv4 (S,G), the only difference in the AMT relay
   discovery process is the use of the in-addr.arpa reverse IP domain
   name, as described in Section 3.5 of [RFC1035], instead of the
   in6.arpa domain name.  For example, if the (S,G) is (198.51.100.12,
   232.252.0.2), the reverse IP FQDN for the AMTRELAY query would be
   "12.100.51.198.in-addr.arpa.".

   Note that the address family of the AMT tunnel is independent of the
   address family for the multicast traffic.

2.3.  Example Deployments



2.3.1.  Example Receiving Networks



2.3.1.1.  Internet Service Provider



   One example of a receiving network is an Internet Service Provider
   (ISP) that offers multicast ingest services to its subscribers,
   illustrated in Figure 3.

   In the example network below, subscribers can join (S,G)s with MLDv2
   or IGMPv3 as described in [RFC4604], and the AMT gateway in this ISP
   can receive and forward multicast traffic from one of the example
   sending networks in Section 2.3.2 by discovering the appropriate AMT
   relays with a DNS lookup for the AMTRELAY RR with the reverse IP of
   the source in the (S,G).

                        Internet
                     ^            ^      Multicast-enabled
                     |            |      Receiving Network
              +------|------------|-------------------------+
              |      |            |                         |
              |  +--------+   +--------+    +=========+     |
              |  | Border |---| Border |    |   AMT   |     |
              |  | Router |   | Router |    | gateway |     |
              |  +--------+   +--------+    +=========+     |
              |      |            |              |          |
              |      +-----+------+-----------+--+          |
              |            |                  |             |
              |      +-------------+    +-------------+     |
              |      | Agg Routers | .. | Agg Routers |     |
              |      +-------------+    +-------------+     |
              |            /     \ \     /         \        |
              | +---------------+         +---------------+ |
              | |Access Systems | ....... |Access Systems | |
              | |(CMTS/OLT/etc.)|         |(CMTS/OLT/etc.)| |
              | +---------------+         +---------------+ |
              |        |                        |           |
              +--------|------------------------|-----------+
                       |                        |
                 +---+-+-+---+---+        +---+-+-+---+---+
                 |   |   |   |   |        |   |   |   |   |
                /-\ /-\ /-\ /-\ /-\      /-\ /-\ /-\ /-\ /-\
                |_| |_| |_| |_| |_|      |_| |_| |_| |_| |_|

                               Subscribers

                      Figure 3: Receiving ISP Example

2.3.1.2.  Small Office



   Another example receiving network is a small branch office that
   regularly accesses some multicast content, illustrated in Figure 4.

   This office has desktop devices that need to receive some multicast
   traffic, so an AMT gateway runs on a LAN with these devices to pull
   traffic in through a non-multicast next hop.

   The office also hosts some mobile devices that have AMT gateway
   instances embedded inside apps in order to receive multicast traffic
   over their non-multicast wireless LAN.  (Note that the "Legacy
   Router" is a simplification that's meant to describe a variety of
   possible conditions; for example, it could be a device providing a
   split-tunnel VPN as described in [RFC7359], deliberately excluding
   multicast traffic for a VPN tunnel, rather than a device that is
   incapable of multicast forwarding.)

                     Internet
                  (non-multicast)
                         ^
                         |                  Office Network
              +----------|----------------------------------+
              |          |                                  |
              |    +---------------+ (Wifi)   Mobile apps   |
              |    | Modem+ | Wifi | - - - -  w/ embedded   |
              |    | Router |  AP  |          AMT gateways  |
              |    +---------------+                        |
              |          |                                  |
              |          |                                  |
              |     +----------------+                      |
              |     | Legacy Router  |                      |
              |     |   (unicast)    |                      |
              |     +----------------+                      |
              |      /        |      \                      |
              |     /         |       \                     |
              | +--------+ +--------+ +--------+=========+  |
              | | Phones | | ConfRm | | Desks  |   AMT   |  |
              | | subnet | | subnet | | subnet | gateway |  |
              | +--------+ +--------+ +--------+=========+  |
              |                                             |
              +---------------------------------------------+

                  Figure 4: Small Office (No Multicast Up)

   By adding an AMT relay to this office network as in Figure 5, it's
   possible to make use of multicast services from the example
   multicast-capable ISP in Section 2.3.1.1.

               Multicast-capable ISP
                         ^
                         |                  Office Network
              +----------|----------------------------------+
              |          |                                  |
              |    +---------------+ (Wifi)   Mobile apps   |
              |    | Modem+ | Wifi | - - - -  w/ embedded   |
              |    | Router |  AP  |          AMT gateways  |
              |    +---------------+                        |
              |          |               +=======+          |
              |          +---Wired LAN---|  AMT  |          |
              |          |               | relay |          |
              |     +----------------+   +=======+          |
              |     | Legacy Router  |                      |
              |     |   (unicast)    |                      |
              |     +----------------+                      |
              |      /        |      \                      |
              |     /         |       \                     |
              | +--------+ +--------+ +--------+=========+  |
              | | Phones | | ConfRm | | Desks  |   AMT   |  |
              | | subnet | | subnet | | subnet | gateway |  |
              | +--------+ +--------+ +--------+=========+  |
              |                                             |
              +---------------------------------------------+

                       Figure 5: Small Office Example

   When multicast-capable networks are chained like this, with a network
   like the one in Figure 5 receiving Internet services from a
   multicast-capable network like the one in Figure 3, it's important
   for AMT gateways to reach the more local AMT relay in order to avoid
   accidentally tunneling multicast traffic from a more distant AMT
   relay with unicast and failing to utilize the multicast transport
   capabilities of the network in Figure 3.

2.3.2.  Example Sending Networks



2.3.2.1.  Sender-Controlled Relays



   When a sender network is also operating AMT relays to distribute
   multicast traffic, as in Figure 6, each address could appear as an
   AMTRELAY RR for the reverse IP of the sender.  Alternately, one or
   more domain names could appear in AMTRELAY RRs, and the AMT relay
   addresses can be discovered by finding A or AAAA records from those
   domain names.

                                          Sender Network
                    +-----------------------------------+
                    |                                   |
                    | +--------+   +=======+  +=======+ |
                    | | Sender |   |  AMT  |  |  AMT  | |
                    | +--------+   | relay |  | relay | |
                    |     |        +=======+  +=======+ |
                    |     |            |          |     |
                    |     +-----+------+----------+     |
                    |           |                       |
                    +-----------|-----------------------+
                                v
                             Internet
                          (non-multicast)

                       Figure 6: Small Office Example

2.3.2.2.  Provider-Controlled Relays



   When an ISP offers a service to transmit outbound multicast traffic
   through a forwarding network, it might also offer AMT relays in order
   to reach receivers without multicast connectivity to the forwarding
   network, as in Figure 7.  In this case, it's recommended that the ISP
   also provide at least one domain name for the AMT relays for use with
   the AMTRELAY RR.

   When the sender wishes to use the relays provided by the ISP for
   forwarding multicast traffic, an AMTRELAY RR should be configured to
   use the domain name provided by the ISP to allow for address
   reassignment of the relays without forcing the sender to reconfigure
   the corresponding AMTRELAY RRs.

                      +--------+
                      | Sender |
                      +---+----+        Multicast-enabled
                          |              Sending Network
              +-----------|-------------------------------+
              |           v                               |
              |    +------------+     +=======+ +=======+ |
              |    | Agg Router |     |  AMT  | |  AMT  | |
              |    +------------+     | relay | | relay | |
              |           |           +=======+ +=======+ |
              |           |               |         |     |
              |     +-----+------+--------+---------+     |
              |     |            |                        |
              | +--------+   +--------+                   |
              | | Border |---| Border |                   |
              | | Router |   | Router |                   |
              | +--------+   +--------+                   |
              +-----|------------|------------------------+
                    |            |
                    v            v
                       Internet
                    (non-multicast)

                       Figure 7: Sending ISP Example

3.  Relay Discovery Operation



3.1.  Optimal Relay Selection



3.1.1.  Overview



   The reverse source IP DNS query of an AMTRELAY RR is a good way for a
   gateway to discover a relay that is known to the sender.

   However, it is *not* necessarily a good way to discover the best
   relay for that gateway to use, because the RR will only provide
   information about relays known to the source.

   If there is an upstream relay in a network that is topologically
   closer to the gateway and is able to receive and forward multicast
   traffic from the sender, that relay is better for the gateway to use
   since more of the network path uses native multicast, allowing more
   chances for packet replication.  But since that relay is not known to
   the sender, it won't be advertised in the sender's reverse IP DNS
   record.  An example network that illustrates this scenario is
   outlined in Figure 5 from Section 2.3.1.2.

   It's only appropriate for an AMT gateway to discover an AMT relay by
   querying an AMTRELAY RR owned by a sender when all of these
   conditions are met:

   1.  The gateway needs to propagate a join of an (S,G) over AMT
       because in the gateway's network, no RPF next hop toward the
       source can propagate a native multicast join of the (S,G);

   2.  The gateway is not already connected to a relay that forwards
       multicast traffic from the source of the (S,G);

   3.  The gateway is not configured to use a particular IP address for
       AMT discovery, or a relay discovered with that IP is not able to
       forward traffic from the source of the (S,G);

   4.  The gateway is not able to find an upstream AMT relay with DNS-
       based Service Discovery (DNS-SD) [RFC6763] using "_amt._udp" as
       the Service section of the queries, or a relay discovered this
       way is not able to forward traffic from the source of the (S,G)
       (as described in Section 3.3.4.1 and 3.3.5); and

   5.  The gateway is not able to find an upstream AMT relay with the
       well-known anycast addresses from Section 7 of [RFC7450].



   When all of the above conditions are met, the gateway has no path
   within its local network that can receive multicast traffic from the
   source IP of the (S,G).

   In this situation, the best way to find a relay that can forward the
   required traffic is to use information that comes from the operator
   of the sender.  When the sender has configured an AMTRELAY RR,
   gateways can use the DRIAD mechanism defined in this document to
   discover the relay information provided by the sender.

   Note that the above conditions are designed to prefer the use of a
   local AMT relay if one can be discovered.  However, note also that
   the network upstream of the locally discovered relay would still need
   to receive traffic from the sender of the (S,G) in order to forward
   it.  Therefore, unless the upstream network contains the sender or
   has a multicast-capable peering with a network that can forward
   traffic from the sender, the upstream network might still use AMT to
   ingest the traffic from a network that can receive traffic from the
   sender.  If this is the case, the upstream AMT gateway could still
   rely on the AMTRELAY RR provided by the sender, even though the
   AMTRELAY RR is not directly used by gateways topologically closer to
   the receivers.  For a concrete example of such a situation, consider
   the network in Figure 5 connected as one of the customers to the
   network in Figure 3.

3.1.2.  Preference Ordering



   This section defines a preference ordering for relay addresses during
   the relay discovery process.  Gateways are encouraged to implement a
   Happy Eyeballs [RFC8305] algorithm to try candidate relays
   concurrently (see Section 3.2), but even gateways that do not
   implement a Happy Eyeballs algorithm SHOULD use this ordering, except
   as noted.

   When establishing an AMT tunnel to forward multicast data, it's very
   important for the discovery process to prioritize network topology
   considerations ahead of address selection considerations in order to
   gain the packet replication benefits from using multicast instead of
   unicast tunneling in the multicast-capable portions of the network
   path.

   The intent of the advice and requirements in this section is to
   describe how a gateway should make use of the concurrency provided by
   a Happy Eyeballs algorithm to reduce the join latency while still
   prioritizing network efficiency considerations over address selection
   considerations.

   Section 4 of [RFC8305] requires a Happy Eyeballs algorithm to sort
   the addresses with the Destination Address Selection defined in
   Section 6 of [RFC6724], but for the above reasons, that requirement
   is superseded in the AMT discovery use case by the following
   considerations:

   *  Prefer Local Relays

      Figure 5 and Section 2.3.1.2 provide a motivating example to
      prefer DNS-SD [RFC6763] for discovery strictly ahead of using the
      AMTRELAY RR controlled by the sender for AMT discovery.

      For this reason, it's RECOMMENDED that AMT gateways by default
      perform service discovery using DNS Service Discovery (DNS-SD)
      [RFC6763] for _amt._udp.<domain> (with <domain> chosen as
      described in Section 11 of [RFC6763]) and use the AMT relays
      discovered that way in preference to AMT relays discoverable via
      the mechanism defined in this document (DRIAD).

   *  Prefer Relays Managed by the Containing Network

      When no local relay is discoverable with DNS-SD, it still may be
      the case that a relay local to the receiver is operated by the
      network providing transit services to the receiver.

      In this case, when the network cannot make the relay discoverable
      via DNS-SD, the network SHOULD use the well-known anycast
      addresses from Section 7 of [RFC7450] to route discovery traffic
      to the relay most appropriate to the receiver's gateway.

      Accordingly, the gateway SHOULD by default discover a relay with
      the well-known AMT anycast addresses as the next-best option after
      DNS-SD when searching for a local relay.

   *  Let Sender Manage Relay Provisioning

      A related motivating example is provided by considering a sender
      whose traffic can be forwarded by relays in a sender-controlled
      network like Figure 6 in Section 2.3.2.1 and by relays in a
      provider-controlled network like Figure 7 in Section 2.3.2.2, with
      different cost and scalability profiles for the different options.

      In this example about the sending-side network, the precedence
      field described in Section 4.2.1 is a critical method of control
      so that senders can provide the appropriate guidance to gateways
      during the discovery process in order to manage load and failover
      scenarios in a manner that operates well with the sender's
      provisioning strategy for horizontal scaling of AMT relays.

      Therefore, after DNS-SD, the precedence from the RR MUST be used
      for sorting preference ahead of the Destination Address Selection
      ordering from Section 6 of [RFC6724] so that only relay IPs with
      the same precedence are directly compared according to the
      Destination Address Selection ordering.

   Accordingly, AMT gateways SHOULD by default prefer relays in this
   order:

   1.  DNS-SD

   2.  Anycast addresses from Section 7 of [RFC7450]

   3.  DRIAD

   This default behavior MAY be overridden by administrative
   configuration where other behavior is more appropriate for the
   gateway within its network.

   Among relay addresses that have an equivalent preference as described
   above, a Happy Eyeballs algorithm for AMT SHOULD use the Destination
   Address Selection defined in Section 6 of [RFC6724].

   Among relay addresses that still have an equivalent preference after
   the above orderings, a gateway SHOULD make a non-deterministic choice
   (such as a pseudorandom selection) for relay preference ordering in
   order to support load balancing by DNS configurations that provide
   many relay options.

   The gateway MAY introduce a bias in the non-deterministic choice
   according to information that indicates expected benefits from
   selecting some relays in preference to others.  Details about the
   structure and collection of this information are out of scope for
   this document but could, for example, be obtained by out-of-band
   methods or from a historical record about network topology, timing
   information, or the response to a probing mechanism.  A gateway in
   possession of such information MAY use it to prefer topologically
   closer relays.

   Within the above constraints, gateways MAY make use of other
   considerations from Section 4 of [RFC8305], such as the address
   family interleaving strategies, to produce a final ordering of
   candidate relay addresses.

   Note also that certain relay addresses might be excluded from
   consideration by the hold-down timers described in Section 3.3.4.1 or
   3.3.5.  These relays constitute "unusable destinations" under Rule 1
   of the Destination Address Selection and are also not part of the
   superseding considerations described above.

   The discovery and connection process for the relay addresses in the
   above described ordering MAY operate in parallel, subject to delays
   prescribed by the Happy Eyeballs requirements described in Section 5
   of [RFC8305] for successively launched concurrent connection
   attempts.

3.1.3.  Connecting to Multiple Relays



   In some deployments, it may be useful for a gateway to connect to
   multiple upstream relays and subscribe to the same traffic in order
   to support an active/active failover model.  A gateway SHOULD NOT be
   configured to do so without guaranteeing that adequate bandwidth is
   available.

   A gateway configured to do this SHOULD still use the same preference-
   ordering logic from Section 3.1.2 for each connection.  (Note that
   this ordering allows for overriding by explicit administrative
   configuration where required.)

3.2.  Happy Eyeballs



3.2.1.  Overview



   Often, multiple choices of relay will exist for a gateway using DRIAD
   for relay discovery.  Happy Eyeballs [RFC8305] provides a widely
   deployed and generalizable strategy for probing multiple possible
   connections in parallel.  Therefore, it is RECOMMENDED that DRIAD-
   capable gateways implement a Happy Eyeballs algorithm to support fast
   discovery of the most preferred available relay by probing multiple
   relays concurrently.

   The parallel discovery logic of a Happy Eyeballs algorithm serves to
   reduce join latency for the initial join of an SSM channel.  This
   section and the preference ordering of relays defined in
   Section 3.1.2 together provide guidance on use of a Happy Eyeballs
   algorithm for the case of establishing AMT connections.

   Note that according to the definition in Section 3.2.3 of this
   document, establishing the connection occurs before sending a
   membership report.  As described in Section 5 of [RFC8305], only one
   of the successful connections will be used, and the others are all
   canceled or ignored.  In the context of an AMT connection, this means
   the gateway will send the membership reports that subscribe to
   traffic only for the chosen connection after the Happy Eyeballs
   algorithm resolves.

3.2.2.  Algorithm Guidelines



   During the "Initiation of asynchronous DNS queries" phase described
   in Section 3 of [RFC8305], a gateway attempts to resolve the domain
   names listed in Section 3.1.  This consists of resolving the SRV
   queries for DNS-SD domains for the AMT service, as well as the
   AMTRELAY query for the reverse IP domain defined in this document.

   Each of the SRV and AMTRELAY responses might contain:

   *  one or more IP addresses (as with type 1 or type 2 AMTRELAY
      responses or when the SRV Additional Data section of the SRV
      response contains the address records for the target, as urged by
      [RFC2782]), or

   *  only domain names (as with type 3 responses from Section 4.2.3 or
      an SRV response without an additional data section).

   When present, IP addresses in the initial response provide resolved
   destination address candidates for the "Sorting of resolved
   destination addresses" phase described in Section 4 of [RFC8305]),
   whereas domain names without IP addresses in the initial response
   result in another set of queries for AAAA and A records, whose
   responses provide the candidate resolved destination addresses.

   Since the SRV or AMTRELAY responses don't have a bound on the count
   of queries that might be generated aside from the bounds imposed by
   the DNS resolver, it's important for the gateway to provide a rate
   limit on the DNS queries.  The DNS query functionality is expected to
   follow ordinary standards and best practices for DNS clients.  A
   gateway MAY use an existing DNS client implementation that does so
   and MAY rely on that client's rate-limiting logic to avoid issuing
   excessive queries.  Otherwise, a gateway MUST provide a rate limit
   for the DNS queries, and its default settings SHOULD NOT permit more
   than 10 queries for any 100-millisecond period (though this MAY be
   overridable by the administrative configuration).

   As the resolved IP addresses arrive, the Happy Eyeballs algorithm
   sorts them according to the requirements and recommendations given in
   Section 3.1.2 and attempts connections with the corresponding relays
   under the algorithm restrictions and guidelines given in [RFC8305]
   for the "Establishment of one connection, which cancels all other
   attempts" phase.  As described in Section 3 of [RFC8305], DNS
   resolution is treated as asynchronous, and connection initiation does
   not wait for lagging DNS responses.

3.2.3.  Connection Definition



   Section 5 of [RFC8305] non-normatively describes a successful
   connection attempt as "generally when the TCP handshake completes".

   There is no normative definition of a connection in the AMT
   specification [RFC7450], and there is no TCP connection involved in
   an AMT tunnel.

   However, the concept of an AMT connection in the context of a Happy
   Eyeballs algorithm is a useful one, and so this section provides the
   following normative definition:

   *  An AMT connection is established successfully when the gateway
      receives from a newly discovered relay a valid Membership Query
      message (Section 5.1.4 of [RFC7450]) that does not have the L flag
      set.

   See Section 3.3.5 of this document for further information about the
   relevance of the L flag to the establishment of a Happy Eyeballs
   connection.  See Section 3.3.4 for an overview of how to respond if
   the connection does not provide multicast connectivity to the source.

   To "cancel" this kind of AMT connection for the Happy Eyeballs
   algorithm, a gateway that has not sent a membership report with a
   subscription would simply stop sending AMT packets for that
   connection.  A gateway only sends a membership report to a connection
   it has chosen as the most preferred available connection.

3.3.  Guidelines for Restarting Discovery



3.3.1.  Overview



   It's expected that gateways deployed in different environments will
   use a variety of heuristics to decide when it's appropriate to
   restart the relay discovery process in order to meet different
   performance goals (for example, to fulfill different kinds of service
   level agreements).

   In general, restarting the discovery process is always safe for the
   gateway and relay during any of the events listed in this section but
   may cause a disruption in the forwarded traffic if the discovery
   process results in choosing a different relay because this changes
   the RPF forwarding tree for the multicast traffic upstream of the
   gateway.  This is likely to result in some dropped or duplicated
   packets from channels actively being tunneled from the old relay to
   the gateway.

   The degree of impact on the traffic from choosing a different relay
   may depend on network conditions between the gateway and the new
   relay, as well as the network conditions and topology between the
   sender and the new relay, as this may cause the relay to propagate a
   new RPF join toward the sender.

   Balancing the expected impact on the tunneled traffic against likely
   or observed problems with an existing connection to the relay is the
   goal of the heuristics that gateways use to determine when to restart
   the discovery process.

   The non-normative advice in this section should be treated as
   guidelines to operators and implementors working with AMT systems
   that can use DRIAD as part of the relay discovery process.

3.3.2.  Updates to Restarting Events



   Section 5.2.3.4.1 of [RFC7450] lists several events that may cause a
   gateway to start or restart the discovery procedure.

   This document provides some updates and recommendations regarding the
   handling of these and similar events.  The first five events are
   copied here and numbered for easier reference, and the remaining four
   events are newly added for consideration in this document:

   1.  When a gateway pseudo-interface is started (enabled).

   2.  When the gateway wishes to report a group subscription when none
       currently exists.

   3.  Before sending the next Request message in a membership update
       cycle.

   4.  After the gateway fails to receive a response to a Request
       message.



   5.  After the gateway receives a Membership Query message with the L
       flag set to 1.

   6.  When the gateway wishes to report an (S,G) subscription with a
       source address that does not currently have other group
       subscriptions.



   7.  When there is a network change detected; for example, when a
       gateway is operating inside an end user device or application and
       the device joins a different network or when the domain portion
       of a DNS-SD domain name changes in response to a DHCP message or
       administrative configuration.

   8.  When substantial loss, persistent congestion, or network overload
       is detected in the stream of AMT packets from a relay.



   9.  When the gateway has reported one or more (S,G) subscriptions but
       no traffic is received from the source for some timeout (see
       Section 3.3.4.1).



   This list is not exhaustive, nor are any of the listed events
   strictly required to always force a restart of the discovery process.

   Note that during event #1, a gateway may use DNS-SD but does not have
   sufficient information to use DRIAD, since no source is known.

3.3.3.  Tunnel Stability



   In general, subscribers to active traffic flows that are being
   forwarded by an AMT gateway are less likely to experience a
   degradation in service (for example, from missing or duplicated
   packets) when the gateway continues using the same relay as long as
   the relay is not overloaded and the network conditions remain stable.

   Therefore, gateways SHOULD avoid performing a full restart of the
   discovery process during routine cases of event #3 (sending a new
   Request message), since it occurs frequently in normal operation.

   However, see Sections 3.3.4, 3.3.6, and 3.3.4.3 for more information
   about exceptional cases when it may be appropriate to use event #3.

3.3.4.  Traffic Health



3.3.4.1.  Absence of Traffic



   If a gateway indicates one or more (S,G) subscriptions in a
   Membership Update message but no traffic for any of the (S,G)s is
   received in a reasonable time, it's appropriate for the gateway to
   restart the discovery process.

   If the gateway restarts the discovery process multiple times
   consecutively for this reason, the timeout period SHOULD be adjusted
   to provide a random exponential back-off.

   The RECOMMENDED timeout is a random value in the range
   [initial_timeout, MIN(initial_timeout * 2^retry_count,
   maximum_timeout)], with a RECOMMENDED initial_timeout of 4 seconds
   and a RECOMMENDED maximum_timeout of 120 seconds (which is the
   recommended minimum NAT mapping timeout described in [RFC4787]).

   Note that the recommended initial_timeout is larger than the initial
   timeout recommended in the similar algorithm from Section 5.2.3.4.3
   of [RFC7450].  This is to provide time for RPF Join propagation in
   the sending network.  Although the timeout values may be
   administratively adjusted to support performance requirements,
   operators are advised to consider the possibility of join propagation
   delays between the sender and the relay when choosing an appropriate
   timeout value.

   Gateways restarting the discovery process because of an absence of
   traffic MUST use a hold-down timer that removes this relay from
   consideration during subsequent rounds of discovery while active.
   The hold-down SHOULD last for no less than 3 minutes and no more than
   10 minutes.

3.3.4.2.  Loss and Congestion



   In some gateway deployments, it is also feasible to monitor the
   health of traffic flows through the gateway -- for example, by
   detecting the rate of packet loss by communicating out of band with
   receivers or monitoring the packets of known protocols with sequence
   numbers.  Where feasible, it's encouraged for gateways to use such
   traffic health information to trigger a restart of the discovery
   process during event #3 (before sending a new Request message).

   However, if a transient network event that affects the tunneled
   multicast stream -- as opposed to an event that affects the tunnel
   connection between the relay and gateway -- occurs, poor health
   detection could be triggered for many gateways simultaneously.  In
   this situation, adding a random delay to avoid synchronized
   rediscovery by many gateways is recommended.

   The span of the random portion of the delay should be no less than 10
   seconds by default but may be administratively configured to support
   different performance requirements.

3.3.4.3.  Ancient Discovery Information



   In most cases, a gateway actively receiving healthy traffic from a
   relay that has not indicated load with the L flag should prefer to
   remain connected to the same relay, as described in Section 3.3.3.

   However, a relay that appears healthy but has been forwarding traffic
   for days or weeks may have an increased chance of becoming unstable.
   Gateways may benefit from restarting the discovery process during
   event #3 (before sending a Request message) after the expiration of a
   long-term timeout on the order of multiple hours or even days in some
   deployments.

   It may be beneficial for such timers to consider the amount of
   traffic currently being forwarded and to give a higher probability of
   restarting discovery during periods with an unusually low data rate
   to reduce the impact on active traffic while still avoiding relying
   on the results of a very old discovery.

   Other issues may also be worth considering as part of this heuristic;
   for example, if the DNS expiry time of the record that was used to
   discover the current relay has not passed, the long-term timer might
   be restarted without restarting the discovery process.

3.3.5.  Relay Loaded or Shutting Down



   The L flag (see Section 5.1.4.4 of [RFC7450]) is the preferred
   mechanism for a relay to signal overloading or a graceful shutdown to
   gateways.

   A gateway that supports handling of the L flag should generally
   restart the discovery process when it processes a Membership Query
   packet with the L flag set.  If an L flag is received while a
   concurrent Happy Eyeballs discovery process is underway for multiple
   candidate relays (Section 3.2), the relay sending the L flag SHOULD
   NOT
be considered for the relay selection.

   It is also RECOMMENDED that gateways avoid choosing a relay that has
   recently sent an L flag, with approximately a 10-minute hold-down.
   Gateways SHOULD treat this hold-down timer in the same way as the
   hold-down in Section 3.3.4.1 so that the relay is removed from
   consideration for subsequent short-term rounds of discovery.

3.3.6.  Relay Discovery Messages vs. Restarting Discovery



   All AMT relays are required by [RFC7450] to support handling of Relay
   Discovery messages (e.g., in Section 5.3.3.2 of [RFC7450]).

   So a gateway with an existing connection to a relay can send a Relay
   Discovery message to the unicast address of that AMT relay.  Under
   stable conditions with an unloaded relay, it's expected that the
   relay will return its own unicast address in the Relay Advertisement
   in response to such a Relay Discovery message.  Since this will not
   result in the gateway changing to another relay unless the relay
   directs the gateway away, this is a reasonable exception to the
   advice against handling event #3 described in Section 3.3.3.

   This behavior is discouraged for gateways that do support the L flag
   to avoid sending unnecessary packets over the network.

   However, gateways that do not support the L flag may be able to avoid
   a disruption in the forwarded traffic by sending such Relay Discovery
   messages regularly.  When a relay is under load or has started a
   graceful shutdown, it may respond with a different relay address,
   which the gateway can use to connect to a different relay.  This kind
   of coordinated handoff will likely result in a smaller disruption to
   the traffic than if the relay simply stops responding to Request
   messages and stops forwarding traffic.

   This style of Relay Discovery message (one sent to the unicast
   address of a relay that's already forwarding traffic to this gateway)
   SHOULD NOT be considered a full restart of the relay discovery
   process.  It is RECOMMENDED that gateways support the L flag, but for
   gateways that do not support the L flag, sending this message during
   event #3 may help mitigate service degradation when relays become
   unstable.

3.3.7.  Independent Discovery per Traffic Source



   Relays discovered via the AMTRELAY RR are source-specific relay
   addresses and may use different pseudo-interfaces from each other and
   from relays discovered via DNS-SD or via a non-source-specific
   address, as described in Section 4.1.2.1 of [RFC7450].

   Restarting the discovery process for one pseudo-interface does not
   require restarting the discovery process for other pseudo-interfaces.
   Gateway heuristics about restarting the discovery process should
   operate independently for different tunnels to relays when responding
   to events that are specific to the different tunnels.

3.4.  DNS Configuration



   Often, an AMT gateway will only have access to the source and group
   IP addresses of the desired traffic and will not know any other name
   for the source of the traffic.  Because of this, typically, the best
   way of looking up AMTRELAY RRs will be by using the source IP address
   as an index into one of the reverse mapping trees (in-addr.arpa for
   IPv4, as described in Section 3.5 of [RFC1035], or ip6.arpa for IPv6,
   as described in Section 2.5 of [RFC3596]).

   Therefore, it is RECOMMENDED that AMTRELAY RRs be added to reverse IP
   zones as appropriate.  AMTRELAY records MAY also appear in other
   zones, since this may be necessary to perform delegation from the
   reverse zones (see, for example, Section 5.2 of [RFC2317]), but the
   use case enabled by this document requires a reverse IP mapping for
   the source from an (S,G) in order to be useful to most AMT gateways.
   This document does not define semantics for the use of AMTRELAY
   records obtained in a way other than by a reverse IP lookup.

   When performing the AMTRELAY RR lookup, any CNAMEs or DNAMEs found
   MUST be followed.  This is necessary to support zone delegation.
   Some examples outlining this need are described in [RFC2317].

   See Sections 4 and 4.3 for a detailed explanation of the contents of
   a DNS zone file.

3.5.  Waiting for DNS Resolution



   DNS query functionality is expected to follow ordinary standards and
   best practices for DNS clients.  A gateway MAY use an existing DNS
   client implementation that does so and MAY rely on that client's
   retry logic to determine the timeouts between retries.

   Otherwise, a gateway MAY resend a DNS query if it does not receive an
   appropriate DNS response within some timeout period.  If the gateway
   retries multiple times, the timeout period SHOULD be adjusted to
   provide a random exponential back-off.

   As with the waiting process for the Relay Advertisement message from
   Section 5.2.3.4.3 of [RFC7450], the RECOMMENDED timeout is a random
   value in the range [initial_timeout, MIN(initial_timeout *
   2^retry_count, maximum_timeout)], with a RECOMMENDED initial_timeout
   of 1 second and a RECOMMENDED maximum_timeout of 120 seconds.

4.  AMTRELAY Resource Record Definition

4.1.  AMTRELAY RRType



   The AMTRELAY RRType has the mnemonic AMTRELAY and type code 260
   (decimal).

   The AMTRELAY RR is class independent.

4.2.  AMTRELAY RData Format



   The AMTRELAY RData consists of an 8-bit precedence field, a 1-bit
   "Discovery Optional" field, a 7-bit type field, and a variable length
   relay field.

    0                   1                   2                   3
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |   precedence  |D|    type     |                               |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+                               +
   ~                            relay                              ~
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

4.2.1.  RData Format - Precedence



   This is an 8-bit precedence for this record.  It is interpreted in
   the same way as the PREFERENCE field described in Section 3.3.9 of
   [RFC1035].

   Relays listed in AMTRELAY records with a lower value for precedence
   are to be attempted first.

4.2.2.  RData Format - Discovery Optional (D-bit)



   The D-bit is a "Discovery Optional" flag.

   If the D-bit is set to 0, a gateway using this RR MUST perform AMT
   relay discovery as described in Section 4.2.1.1 of [RFC7450] rather
   than directly sending an AMT Request message to the relay.

   That is, the gateway MUST receive an AMT Relay Advertisement message
   (Section 5.1.2 of [RFC7450]) for an address before sending an AMT
   Request message (Section 5.1.3 of [RFC7450]) to that address.  Before
   receiving the Relay Advertisement message, this record has only
   indicated that the address can be used for AMT relay discovery, not
   for a Request message.  This is necessary for devices that are not
   fully functional AMT relays but rather load balancers or brokers, as
   mentioned in Section 4.2.1.1 of [RFC7450].

   If the D-bit is set to 1, the gateway MAY send an AMT Request message
   directly to the discovered relay address without first sending an AMT
   Discovery message.

   This bit should be set according to advice from the AMT relay
   operator.  The D-bit MUST be set to zero when no information is
   available from the AMT relay operator about its suitability.

4.2.3.  RData Format - Type



   The type field indicates the format of the information that is stored
   in the relay field.

   The following values are defined:

   *  type = 0: The relay field is empty (0 bytes).

   *  type = 1: The relay field contains a 4-octet IPv4 address.

   *  type = 2: The relay field contains a 16-octet IPv6 address.

   *  type = 3: The relay field contains a wire-encoded domain name.
      The wire-encoded format is self-describing, so the length is
      implicit.  The domain name MUST NOT be compressed (see Section 3.3
      of [RFC1035] and Section 4 of [RFC3597]).

   RRs with an undefined value in the Type field SHOULD NOT be
   considered by receiving gateways for AMT relay discovery.

4.2.4.  RData Format - Relay



   The relay field is the address or domain name of the AMT relay.  It
   is formatted according to the type field.

   When the type field is 0, the length of the relay field is 0, and it
   indicates that no AMT relay should be used for multicast traffic from
   this source.

   When the type field is 1, the length of the relay field is 4 octets,
   and a 32-bit IPv4 address is present.  This is an IPv4 address as
   described in Section 3.4.1 of [RFC1035].  This is a 32-bit number in
   network byte order.

   When the type field is 2, the length of the relay field is 16 octets,
   and a 128-bit IPv6 address is present.  This is an IPv6 address as
   described in Section 2.2 of [RFC3596].  This is a 128-bit number in
   network byte order.

   When the type field is 3, the relay field is a normal wire-encoded
   domain name, as described in Section 3.3 of [RFC1035].  For the
   reasons given in Section 4 of [RFC3597], compression MUST NOT be
   used.

   For a type 3 record, the D-bit and preference fields carry over to
   all A or AAAA records for the domain name.  There is no difference in
   the result of the discovery process when it's obtained by type 1 or
   type 2 AMTRELAY records with identical D-bit and preference fields
   vs. when the result is obtained by a type 3 AMTRELAY record that
   resolves to the same set of IPv4 and IPv6 addresses via A and AAAA
   lookups.

4.3.  AMTRELAY Record Presentation Format



4.3.1.  Representation of AMTRELAY RRs



   AMTRELAY RRs may appear in a zone data master file.  The precedence,
   D-bit, relay type, and relay fields are REQUIRED.

   If the relay type field is 0, the relay field MUST be ".".

   The presentation for the record is as follows:

     IN AMTRELAY precedence D-bit type relay

4.3.2.  Examples



   In a DNS authoritative nameserver that understands the AMTRELAY type,
   the zone might contain a set of entries like this:

       $ORIGIN 100.51.198.in-addr.arpa.
       12     IN AMTRELAY  10 0 1 203.0.113.15
       12     IN AMTRELAY  10 0 2 2001:db8::15
       12     IN AMTRELAY 128 1 3 amtrelays.example.com.

   This configuration advertises an IPv4 discovery address, an IPv6
   discovery address, and a domain name for AMT relays that can receive
   traffic from the source 198.51.100.12.  The IPv4 and IPv6 addresses
   are configured with a D-bit of 0 (meaning discovery is mandatory, as
   described in Section 4.2.2) and a precedence 10 (meaning they're
   preferred ahead of the last entry, which has precedence 128).

   For zone files in name servers that don't support the AMTRELAY RRType
   natively, it's possible to use the format for unknown RR types, as
   described in [RFC3597].  This approach would replace the AMTRELAY
   entries in the example above with the entries below:

       10   IN TYPE260  \# (
              6  ; length
              0a ; precedence=10
              01 ; D=0, relay type=1, an IPv4 address
              cb00710f ) ; 203.0.113.15
       10   IN TYPE260  \# (
              18 ; length
              0a ; precedence=10
              02 ; D=0, relay type=2, an IPv6 address
              20010db800000000000000000000000f ) ; 2001:db8::15
       10   IN TYPE260  \# (
              24 ; length
              80 ; precedence=128
              83 ; D=1, relay type=3, a wire-encoded domain name
            09616d7472656c617973076578616d706c6503636f6d ) ; domain name

   See Appendix A for more details.

5.  IANA Considerations

   This document updates the DNS "Resource Record (RR) TYPEs" registry
   by assigning type 260 to the AMTRELAY record.

   This document creates a new registry named "AMTRELAY Resource Record
   Parameters" with a subregistry for the "Relay Type Field".  The
   initial values in the subregistry are:

   +-------+---------------------------------------+
   | Value | Description                           |
   +=======+=======================================+
   | 0     | No relay is present                   |
   +-------+---------------------------------------+
   | 1     | A 4-byte IPv4 address is present      |
   +-------+---------------------------------------+
   | 2     | A 16-byte IPv6 address is present     |
   +-------+---------------------------------------+
   | 3     | A wire-encoded domain name is present |
   +-------+---------------------------------------+
   | 4-255 | Unassigned                            |
   +-------+---------------------------------------+

      Table 2: Initial Contents of the "Relay Type
                    Field" Registry

   Values 0, 1, 2, and 3 are further explained in Sections 4.2.3 and
   4.2.4.  Relay type numbers 4 through 255 can be assigned with a
   policy of Specification Required (as described in [RFC8126]).

6.  Security Considerations

6.1.  Use of AMT



   This document defines a mechanism that enables a more widespread and
   automated use of AMT, even without access to a multicast backbone.
   Operators of networks and applications that include a DRIAD-capable
   AMT gateway are advised to carefully consider the security
   considerations in Section 6 of [RFC7450].

   AMT gateway operators also are encouraged to take appropriate steps
   to ensure the integrity of the data received via AMT, for example, by
   the opportunistic use of IPsec [RFC4301] to secure traffic received
   from AMT relays when IPSECKEY records [RFC4025] are available or when
   a trust relationship with the AMT relays can be otherwise established
   and secured.

   Note that AMT does not itself provide any integrity protection for
   Multicast Data packets (Section 5.1.6 of [RFC7450]), so absent
   protections like those mentioned above, even an off-path attacker who
   discovers the gateway IP, the relay IP, and the relay source port for
   an active AMT connection can inject multicast data packets for a
   joined (S,G) into the data stream if he can get data packets
   delivered to the gateway IP that spoof the relay as the source.

6.2.  Record-Spoofing



   The AMTRELAY resource record contains information that SHOULD be
   communicated to the DNS client without being modified.  The method
   used to ensure the result was unmodified is up to the client.

   There must be a trust relationship between the end consumer of this
   resource record and the DNS server.  This relationship may be end-to-
   end DNSSEC validation or a secure connection to a trusted DNS server
   that provides end-to-end safety to prevent record-spoofing of the
   response from the trusted server.  The connection to the trusted
   server can use any secure channel, such as with a TSIG [RFC2845] or
   SIG(0) [RFC2931] channel, a secure local channel on the host, DNS
   over TLS [RFC7858], DNS over HTTPS [RFC8484], or some other mechanism
   that provides authentication of the RR.

   If an AMT gateway accepts a maliciously crafted AMTRELAY record, the
   result could be a Denial of Service or receivers processing multicast
   traffic from a source under the attacker's control.

6.3.  Congestion



   Multicast traffic, particularly interdomain multicast traffic,
   carries some congestion risks, as described in Section 4 of
   [RFC8085].

   Application implementors and network operators that use AMT gateways
   are advised to take precautions, including monitoring of application
   traffic behavior, traffic authentication at ingest, rate-limiting of
   multicast traffic, and the use of circuit-breaker techniques such as
   those described in Section 3.1.10 of [RFC8085] and similar
   protections at the network level in order to ensure network health in
   the event of misconfiguration, poorly written applications that don't
   follow UDP congestion control principles, or a deliberate attack.

   Section 4.1.4.2 of [RFC7450] and Section 6.1 of [RFC7450] provide
   some further considerations and advice about mitigating congestion
   risk.

7.  References



7.1.  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>.

   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
              Requirement Levels", BCP 14, RFC 2119,
              DOI 10.17487/RFC2119, March 1997,
              <https://www.rfc-editor.org/info/rfc2119>.

   [RFC2181]  Elz, R. and R. Bush, "Clarifications to the DNS
              Specification", RFC 2181, DOI 10.17487/RFC2181, July 1997,
              <https://www.rfc-editor.org/info/rfc2181>.

   [RFC2782]  Gulbrandsen, A., Vixie, P., and L. Esibov, "A DNS RR for
              specifying the location of services (DNS SRV)", RFC 2782,
              DOI 10.17487/RFC2782, February 2000,
              <https://www.rfc-editor.org/info/rfc2782>.

   [RFC3376]  Cain, B., Deering, S., Kouvelas, I., Fenner, B., and A.
              Thyagarajan, "Internet Group Management Protocol, Version
              3", RFC 3376, DOI 10.17487/RFC3376, October 2002,
              <https://www.rfc-editor.org/info/rfc3376>.

   [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>.

   [RFC3597]  Gustafsson, A., "Handling of Unknown DNS Resource Record
              (RR) Types", RFC 3597, DOI 10.17487/RFC3597, September
              2003, <https://www.rfc-editor.org/info/rfc3597>.

   [RFC3810]  Vida, R., Ed. and L. Costa, Ed., "Multicast Listener
              Discovery Version 2 (MLDv2) for IPv6", RFC 3810,
              DOI 10.17487/RFC3810, June 2004,
              <https://www.rfc-editor.org/info/rfc3810>.

   [RFC4604]  Holbrook, H., Cain, B., and B. Haberman, "Using Internet
              Group Management Protocol Version 3 (IGMPv3) and Multicast
              Listener Discovery Protocol Version 2 (MLDv2) for Source-
              Specific Multicast", RFC 4604, DOI 10.17487/RFC4604,
              August 2006, <https://www.rfc-editor.org/info/rfc4604>.

   [RFC4607]  Holbrook, H. and B. Cain, "Source-Specific Multicast for
              IP", RFC 4607, DOI 10.17487/RFC4607, August 2006,
              <https://www.rfc-editor.org/info/rfc4607>.

   [RFC6724]  Thaler, D., Ed., Draves, R., Matsumoto, A., and T. Chown,
              "Default Address Selection for Internet Protocol Version 6
              (IPv6)", RFC 6724, DOI 10.17487/RFC6724, September 2012,
              <https://www.rfc-editor.org/info/rfc6724>.

   [RFC6763]  Cheshire, S. and M. Krochmal, "DNS-Based Service
              Discovery", RFC 6763, DOI 10.17487/RFC6763, February 2013,
              <https://www.rfc-editor.org/info/rfc6763>.

   [RFC7450]  Bumgardner, G., "Automatic Multicast Tunneling", RFC 7450,
              DOI 10.17487/RFC7450, February 2015,
              <https://www.rfc-editor.org/info/rfc7450>.

   [RFC8085]  Eggert, L., Fairhurst, G., and G. Shepherd, "UDP Usage
              Guidelines", BCP 145, RFC 8085, DOI 10.17487/RFC8085,
              March 2017, <https://www.rfc-editor.org/info/rfc8085>.

   [RFC8174]  Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC
              2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174,
              May 2017, <https://www.rfc-editor.org/info/rfc8174>.

   [RFC8305]  Schinazi, D. and T. Pauly, "Happy Eyeballs Version 2:
              Better Connectivity Using Concurrency", RFC 8305,
              DOI 10.17487/RFC8305, December 2017,
              <https://www.rfc-editor.org/info/rfc8305>.

   [RFC8499]  Hoffman, P., Sullivan, A., and K. Fujiwara, "DNS
              Terminology", BCP 219, RFC 8499, DOI 10.17487/RFC8499,
              January 2019, <https://www.rfc-editor.org/info/rfc8499>.

7.2.  Informative References



   [RFC2317]  Eidnes, H., de Groot, G., and P. Vixie, "Classless IN-
              ADDR.ARPA delegation", BCP 20, RFC 2317,
              DOI 10.17487/RFC2317, March 1998,
              <https://www.rfc-editor.org/info/rfc2317>.

   [RFC2845]  Vixie, P., Gudmundsson, O., Eastlake 3rd, D., and B.
              Wellington, "Secret Key Transaction Authentication for DNS
              (TSIG)", RFC 2845, DOI 10.17487/RFC2845, May 2000,
              <https://www.rfc-editor.org/info/rfc2845>.

   [RFC2931]  Eastlake 3rd, D., "DNS Request and Transaction Signatures
              ( SIG(0)s )", RFC 2931, DOI 10.17487/RFC2931, September
              2000, <https://www.rfc-editor.org/info/rfc2931>.

   [RFC3550]  Schulzrinne, H., Casner, S., Frederick, R., and V.
              Jacobson, "RTP: A Transport Protocol for Real-Time
              Applications", STD 64, RFC 3550, DOI 10.17487/RFC3550,
              July 2003, <https://www.rfc-editor.org/info/rfc3550>.

   [RFC4025]  Richardson, M., "A Method for Storing IPsec Keying
              Material in DNS", RFC 4025, DOI 10.17487/RFC4025, March
              2005, <https://www.rfc-editor.org/info/rfc4025>.

   [RFC4301]  Kent, S. and K. Seo, "Security Architecture for the
              Internet Protocol", RFC 4301, DOI 10.17487/RFC4301,
              December 2005, <https://www.rfc-editor.org/info/rfc4301>.

   [RFC4787]  Audet, F., Ed. and C. Jennings, "Network Address
              Translation (NAT) Behavioral Requirements for Unicast
              UDP", BCP 127, RFC 4787, DOI 10.17487/RFC4787, January
              2007, <https://www.rfc-editor.org/info/rfc4787>.

   [RFC5110]  Savola, P., "Overview of the Internet Multicast Routing
              Architecture", RFC 5110, DOI 10.17487/RFC5110, January
              2008, <https://www.rfc-editor.org/info/rfc5110>.

   [RFC6726]  Paila, T., Walsh, R., Luby, M., Roca, V., and R. Lehtonen,
              "FLUTE - File Delivery over Unidirectional Transport",
              RFC 6726, DOI 10.17487/RFC6726, November 2012,
              <https://www.rfc-editor.org/info/rfc6726>.

   [RFC7359]  Gont, F., "Layer 3 Virtual Private Network (VPN) Tunnel
              Traffic Leakages in Dual-Stack Hosts/Networks", RFC 7359,
              DOI 10.17487/RFC7359, August 2014,
              <https://www.rfc-editor.org/info/rfc7359>.

   [RFC7761]  Fenner, B., Handley, M., Holbrook, H., Kouvelas, I.,
              Parekh, R., Zhang, Z., and L. Zheng, "Protocol Independent
              Multicast - Sparse Mode (PIM-SM): Protocol Specification
              (Revised)", STD 83, RFC 7761, DOI 10.17487/RFC7761, March
              2016, <https://www.rfc-editor.org/info/rfc7761>.

   [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>.

   [RFC8126]  Cotton, M., Leiba, B., and T. Narten, "Guidelines for
              Writing an IANA Considerations Section in RFCs", BCP 26,
              RFC 8126, DOI 10.17487/RFC8126, June 2017,
              <https://www.rfc-editor.org/info/rfc8126>.

   [RFC8313]  Tarapore, P., Ed., Sayko, R., Shepherd, G., Eckert, T.,
              Ed., and R. Krishnan, "Use of Multicast across Inter-
              domain Peering Points", BCP 213, RFC 8313,
              DOI 10.17487/RFC8313, January 2018,
              <https://www.rfc-editor.org/info/rfc8313>.

   [RFC8484]  Hoffman, P. and P. McManus, "DNS Queries over HTTPS
              (DoH)", RFC 8484, DOI 10.17487/RFC8484, October 2018,
              <https://www.rfc-editor.org/info/rfc8484>.

Appendix A.  Unknown RRType Construction



   In a DNS resolver that understands the AMTRELAY type, the zone file
   might contain this line:

     IN AMTRELAY 128 0 3 amtrelays.example.com.

   In order to translate this example to appear as an unknown RRType as
   defined in [RFC3597], one could run the following program:

   <CODE BEGINS>
     $ cat translate.py
     #!/usr/bin/env python3
     import sys
     name=sys.argv[1]
     wire=''
     for dn in name.split('.'):
       if len(dn) > 0:
         wire += ('%02x' % len(dn))
         wire += (''.join('%02x'%ord(x) for x in dn))
     print(len(wire)//2) + 2
     print(wire)

     $ ./translate.py amtrelays.example.com
     24
     09616d7472656c617973076578616d706c6503636f6d
   <CODE ENDS>

   The length of the RData and the hex string for the domain name
   "amtrelays.example.com" are the outputs of this program.

   The length of the wire-encoded domain name is 22, so 2 was added to
   that value (1 for the precedence field and 1 for the combined D-bit
   and relay type fields) to get the full length 24 of the RData.  For
   the 2 octets ahead of the domain name, we encode the precedence,
   D-bit, and relay type fields, as described in Section 4.

   This results in a zone file entry like this:

     IN TYPE260  \# ( 24 ; length
             80 ; precedence = 128
             03 ; D-bit=0, relay type=3 (wire-encoded domain name)
           09616d7472656c617973076578616d706c6503636f6d ) ; domain name

Acknowledgements



   This specification was inspired by the previous work of Doug Nortz,
   Robert Sayko, David Segelstein, and Percy Tarapore, presented in the
   MBONED Working Group at IETF 93.

   Thanks to Jeff Goldsmith, Toerless Eckert, Mikael Abrahamsson, Lenny
   Giuliano, Mark Andrews, Sandy Zheng, Kyle Rose, Ben Kaduk, Bill
   Atwood, Tim Chown, Christian Worm Mortensen, Warren Kumari, Dan
   Romanescu, Bernard Aboba, Carlos Pignataro, Niclas Comstedt, Mirja
   Kühlewind, Henning Rogge, Eric Vyncke, Barry Lieba, Roman Danyliw,
   Alissa Cooper, Suresh Krishnan, Adam Roach, and Daniel Franke for
   their very helpful reviews and comments.

Author's Address



   Jake Holland
   Akamai Technologies, Inc.
   150 Broadway
   Cambridge, MA 02144
   United States of America

   Email: jakeholland.net@gmail.com