RFC 8801

Internet Engineering Task Force (IETF)                        P. Pfister
Request for Comments: 8801                                     É. Vyncke
Category: Standards Track                                          Cisco
ISSN: 2070-1721                                                 T. Pauly
                                                              Apple Inc.
                                                             D. Schinazi
                                                              Google LLC
                                                                 W. Shao
                                                               July 2020

             Discovering Provisioning Domain Names and Data


   Provisioning Domains (PvDs) are defined as consistent sets of network
   configuration information.  PvDs allows hosts to manage connections
   to multiple networks and interfaces simultaneously, such as when a
   home router provides connectivity through both a broadband and
   cellular network provider.

   This document defines a mechanism for explicitly identifying PvDs
   through a Router Advertisement (RA) option.  This RA option announces
   a PvD identifier, which hosts can compare to differentiate between
   PvDs.  The option can directly carry some information about a PvD and
   can optionally point to PvD Additional Information that can be
   retrieved using HTTP over TLS.

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

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.  Specification of Requirements
   2.  Terminology
   3.  Provisioning Domain Identification Using Router Advertisements
     3.1.  PvD Option for Router Advertisements
     3.2.  Router Behavior
     3.3.  Non-PvD-Aware Host Behavior
     3.4.  PvD-Aware Host Behavior
       3.4.1.  DHCPv6 Configuration Association
       3.4.2.  DHCPv4 Configuration Association
       3.4.3.  Connection Sharing by the Host
       3.4.4.  Usage of DNS Servers
   4.  Provisioning Domain Additional Information
     4.1.  Retrieving the PvD Additional Information
     4.2.  Operational Consideration to Providing the PvD Additional
     4.3.  PvD Additional Information Format
       4.3.1.  Example
     4.4.  Detecting Misconfiguration and Misuse
   5.  Operational Considerations
     5.1.  Exposing Extra RA Options to PvD-Aware Hosts
     5.2.  Different RAs for PvD-Aware and Non-PvD-Aware Hosts
     5.3.  Enabling Multihoming for PvD-Aware Hosts
     5.4.  Providing Additional Information to PvD-Aware Hosts
   6.  Security Considerations
   7.  Privacy Considerations
   8.  IANA Considerations
     8.1.  Change to IPv6 Neighbor Discovery Option Formats Registry
     8.2.  New Entry in the Well-Known URIs Registry
     8.3.  New Additional Information PvD Keys Registry
     8.4.  New PvD Option Flags Registry
     8.5.  PvD JSON Media Type Registration
   9.  References
     9.1.  Normative References
     9.2.  Informative References

   Authors' Addresses

1.  Introduction

   Provisioning Domains (PvDs) are defined in [RFC7556] as consistent
   sets of network configuration information.  This information includes
   properties that are traditionally associated with a single networking
   interface, such as source addresses, DNS configuration, proxy
   configuration, and gateway addresses.

   Clients that are aware of PvDs can take advantage of multiple network
   interfaces simultaneously.  This enables using two PvDs in parallel
   for separate connections or for multi-path transports.

   While most PvDs today are discovered implicitly (such as by receiving
   information via Router Advertisements from a router on a network that
   a client host directly connects to), [RFC7556] also defines the
   notion of Explicit PvDs.  IPsec Virtual Private Networks are
   considered Explicit PvDs, but Explicit PvDs can also be discovered
   via the local network router.  Discovering Explicit PvDs allows two
   key advancements in managing multiple PvDs:

   1.  The ability to discover and use multiple PvDs on a single
       interface, such as when a local router can provide connectivity
       to two different Internet Service Providers.

   2.  The ability to associate Additional Information about PvDs to
       describe the properties of the network.

   While [RFC7556] defines the concept of Explicit PvDs, it does not
   define the mechanism for discovering multiple Explicit PvDs on a
   single network and their Additional Information.

   This document specifies a way to identify PvDs with Fully Qualified
   Domain Names (FQDNs), called PvD IDs.  Those identifiers are
   advertised in a new Router Advertisement (RA) [RFC4861] option called
   the PvD Option, which, when present, associates the PvD ID with all
   the information present in the Router Advertisement as well as any
   configuration object, such as addresses, derived from it.  The PvD
   Option may also contain a set of other RA options, along with an
   optional inner Router Advertisement message header.  These options
   and optional inner header are only visible to 'PvD-aware' hosts,
   allowing such hosts to have a specialized view of the network

   Since PvD IDs are used to identify different ways to access the
   Internet, multiple PvDs (with different PvD IDs) can be provisioned
   on a single host interface.  Similarly, the same PvD ID could be used
   on different interfaces of a host in order to inform that those PvDs
   ultimately provide equivalent services.

   This document also introduces a mechanism for hosts to retrieve
   optional Additional Information related to a specific PvD by means of
   an HTTP-over-TLS query using a URI derived from the PvD ID.  The
   retrieved JSON object contains Additional Information that would
   typically be considered too large to be directly included in the
   Router Advertisement but might be considered useful to the
   applications, or even sometimes users, when choosing which PvD should
   be used.

   For example, if Alice has both a cellular network provider and a
   broadband provider in her home, her PvD-aware devices and
   applications would be aware of both available uplinks.  These
   applications could fail-over between these networks or run
   connections over both (potentially using multi-path transports).
   Applications could also select specific uplinks based on the
   properties of the network; for example, if the cellular network
   provides free high-quality video streaming, a video-streaming
   application could select that network while most of the other traffic
   on Alice's device uses the broadband provider.

1.1.  Specification of Requirements

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "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.  Terminology

   This document uses the following terminology:

   Provisioning Domain (PvD):  A set of network configuration
      information; for more information, see [RFC7556].

   PvD ID:  A Fully Qualified Domain Name (FQDN) used to identify a PvD.

   Explicit PvD:  A PvD uniquely identified with a PvD ID.  For more
      information, see [RFC7556].

   Implicit PvD:  A PvD that, in the absence of a PvD ID, is identified
      by the host interface to which it is attached and the address of
      the advertising router.  See also [RFC7556].

   PvD-aware host:  A host that supports the association of network
      configuration information into PvDs and the use of these PvDs as
      described in this document.  Also named "PvD-aware node" in

3.  Provisioning Domain Identification Using Router Advertisements

   Explicit PvDs are identified by a PvD ID.  The PvD ID is a Fully
   Qualified Domain Name (FQDN) that identifies the network operator.
   Network operators MUST use names that they own or manage to avoid
   naming conflicts.  The same PvD ID MAY be used in several access
   networks when they ultimately provide identical services (e.g., in
   all home networks subscribed to the same service); else, the PvD ID
   MUST be different to follow Section 2.4 of [RFC7556].

3.1.  PvD Option for Router Advertisements

   This document introduces a Router Advertisement (RA) option called
   the PvD Option.  It is used to convey the FQDN identifying a given
   PvD (see Figure 1), bind the PvD ID with configuration information
   received over DHCPv4 (see Section 3.4.2), enable the use of HTTP over
   TLS to retrieve the PvD Additional Information JSON object (see
   Section 4), as well as contain any other RA options that would
   otherwise be valid in the RA.

    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
   |     Type      |    Length     |H|L|R|     Reserved    | Delay |
   |       Sequence Number         |                             ...
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+                             ...
   ...                         PvD ID FQDN                       ...
   ...             +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   ...             |                  Padding                      |
   |                                                             ...
   ...            Router Advertisement message header            ...
   ...             (Only present when R-flag is set)             ...
   ...                                                             |
   |   Options ...

                        Figure 1: PvD Option Format

   Type:  (8 bits) Set to 21.

   Length:  (8 bits) The length of the option in units of 8 octets,
      including the Type and Length fields, the Router Advertisement
      message header, if any, as well as the RA options that are
      included within the PvD Option.

   H-flag:  (1 bit) 'HTTP' flag stating whether some PvD Additional
      Information is made available through HTTP over TLS, as described
      in Section 4.

   L-flag:  (1 bit) 'Legacy' flag stating whether the PvD is associated
      with IPv4 information assigned using DHCPv4 (see Section 3.4.2).

   R-flag:  (1 bit) 'Router Advertisement' flag stating whether the PvD
      Option header is followed (right after padding to the next 64-bit
      boundary) by a Router Advertisement message header (see
      Section 4.2 of [RFC4861]).  The usage of the inner message header
      is described in Section 3.4.

   Reserved:  (9 bits) Reserved for later use.  It MUST be set to zero
      by the sender and ignored by the receiver.

   Delay:  (4 bits) Unsigned integer used to delay HTTP GET queries from
      hosts by a randomized backoff (see Section 4.1).  If the H-flag is
      not set, senders SHOULD set the delay to zero, and receivers
      SHOULD ignore the value.

   Sequence Number:  (16 bits) Sequence number for the PvD Additional
      Information, as described in Section 4.  If the H-flag is not set,
      senders SHOULD set the Sequence Number to zero, and receivers
      SHOULD ignore the value.

   PvD ID FQDN:  The FQDN used as PvD ID encoded in DNS format, as
      described in Section 3.1 of [RFC1035].  Domain name compression as
      described in Section 4.1.4 of [RFC1035] MUST NOT be used.

   Padding:  Zero or more padding octets to the next 8-octet boundary
      (see Section 4.6 of [RFC4861]).  It MUST be set to zero by the
      sender and ignored by the receiver.

   RA message header:  (16 octets) When the R-flag is set, a full Router
      Advertisement message header as specified in [RFC4861].  The
      sender MUST set the Type field to 134 (the value for "Router
      Advertisement") and set the Code field to 0.  Receivers MUST
      ignore both of these fields.  The Checksum field MUST be set to 0
      by the sender; non-zero checksums MUST be ignored by the receiver
      without causing the processing of the message to fail.  All other
      fields are to be set and parsed as specified in [RFC4861] or any
      updating documents.

   Options:  Zero or more RA options that would otherwise be valid as
      part of the Router Advertisement main body but are instead
      included in the PvD Option so as to be ignored by hosts that are
      not PvD aware.

   Figure 2 shows an example of a PvD Option with "example.org" as the
   PvD ID FQDN and includes both a Recursive DNS Server (RDNSS) option
   and a Prefix Information Option.  It has a Sequence Number of 123 and
   indicates the presence of PvD Additional Information that is expected
   to be fetched with a delay factor of 1.

    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
   | Type: 21      |  Length: 12   |1|0|0|     Reserved    |Delay:1|
   |       Seq number: 123         |      7        |       e       |
   |      x        |       a       |      m        |       p       |
   |      l        |       e       |      3        |       o       |
   |      r        |       g       |      0        |   0 (padding) |
   |   0 (padding) |  0 (padding)  |   0 (padding) |   0 (padding) |
   |  RDNSS option (RFC 8106) length: 5                          ...
   ...                                                           ...
   ...                                                             |
   | Prefix Information Option (RFC 4861) length: 4              ...
   ...                                                             |
   ...                                                             |

                        Figure 2: Example PvD Option

3.2.  Router Behavior

   A router MAY send RAs containing one PvD Option but MUST NOT include
   more than one PvD Option in each RA.  The PvD Option MUST NOT contain
   further PvD Options.

   The PvD Option MAY contain zero, one, or more RA options that would
   otherwise be valid as part of the same RA.  Such options are
   processed by PvD-aware hosts and ignored by other hosts as per
   Section 4.2 of [RFC4861].

   In order to provide multiple different PvDs, a router MUST send
   multiple RAs.  RAs sent from different link-local source addresses
   establish distinct Implicit PvDs in the absence of a PvD Option.
   Explicit PvDs MAY share link-local source addresses with an Implicit
   PvD and any number of other Explicit PvDs.

   In other words, different Explicit PvDs MAY be advertised with RAs
   using the same link-local source address, but different Implicit
   PvDs, advertised by different RAs, MUST use different link-local
   addresses because these Implicit PvDs are identified by the source
   addresses of the RAs.  If a link-local address on the router is
   changed, then any new RA will be interpreted as a different Implicit
   PvD by PvD-aware hosts.

   As specified in [RFC4861] and [RFC6980], when the set of options
   causes the size of an advertisement to exceed the link MTU, multiple
   router advertisements MUST be sent to avoid fragmentation, each
   containing a subset of the options.  In such cases, the PvD Option
   header (i.e., all fields except the Options field) MUST be repeated
   in all the transmitted RAs.  The options within the Options field MAY
   be transmitted only once, included in one of the transmitted PvD

3.3.  Non-PvD-Aware Host Behavior

   As the PvD Option has a new option code, non-PvD-aware hosts will
   simply ignore the PvD Option and all the options it contains (see
   Section 4.2 of [RFC4861]).  This ensures the backward compatibility
   required in Section 3.3 of [RFC7556].  This behavior allows for a
   mixed-mode network where a mix of PvD-aware and non-PvD-aware hosts

3.4.  PvD-Aware Host Behavior

   Hosts MUST associate received RAs and included configuration
   information (e.g., Router Valid Lifetime, Prefix Information
   [RFC4861], Recursive DNS Server [RFC8106], and Routing Information
   [RFC4191] options) with the Explicit PvD identified by the first PvD
   Option present in the received RA, if any, or with the Implicit PvD
   identified by the host interface and the source address of the
   received RA otherwise.  If an RA message header is present both
   within the PvD Option and outside it, the header within the PvD
   Option takes precedence.

   In case multiple PvD Options are found in a given RA, hosts MUST
   ignore all but the first PvD Option.

   If a host receives PvD Options flags that it does not recognize
   (currently in the Reserved field), it MUST ignore these flags.

   Similarly, hosts MUST associate all network configuration objects
   (e.g., default routers, addresses, more specific routes, and DNS
   Recursive Resolvers) with the PvD associated with the RA that
   provisioned the object.  For example, addresses that are generated
   using a received Prefix Information Option (PIO) are associated with
   the PvD of the last received RA that included the given PIO.

   PvD IDs MUST be compared in a case-insensitive manner as defined by
   [RFC4343].  For example, "pvd.example.com." or "PvD.Example.coM."
   would refer to the same PvD.

   While performing PvD-specific operations such as resolving names,
   executing the default address selection algorithm [RFC6724], or
   executing the default router selection algorithm when forwarding
   packets [RFC4861] [RFC4191] [RFC8028], hosts and applications MAY
   consider only the configuration associated with any non-empty subset
   of PvDs.  For example, a host MAY associate a given process with a
   specific PvD, or a specific set of PvDs, while associating another
   process with another PvD.  A PvD-aware application might also be able
   to select, on a per-connection basis, which PvDs should be used.  In
   particular, constrained devices such as small battery-operated
   devices (e.g., Internet of Things (IoT)) or devices with limited CPU
   or memory resources may purposefully use a single PvD while ignoring
   some received RAs containing different PvD IDs.

   The way an application expresses its desire to use a given PvD, or a
   set of PvDs, and the way this selection is enforced are out of the
   scope of this document.  Useful insights about these considerations
   can be found in [MPVD-API].

3.4.1.  DHCPv6 Configuration Association

   When a host retrieves stateless configuration elements using DHCPv6
   (e.g., DNS recursive resolvers or DNS domain search lists [RFC3646]),
   they MUST be associated with all the Explicit and Implicit PvDs
   received on the same interface and contained in an RA with the O-flag
   set [RFC4861].

   When a host retrieves stateful assignments using DHCPv6, such
   assignments MUST be associated with the received PvD that was
   received with RAs with the M-flag set and including a matching PIO.
   A PIO is considered to match a DHCPv6 assignment when the IPv6 prefix
   from the PIO includes the assignment from DHCPv6.  For example, if a
   PvD's associated PIO defines the prefix "2001:db8:cafe::/64", a
   DHCPv6 IA_NA message that assigns the address
   "2001:db8:cafe::1234:4567" would be considered to match.

   In cases where an address would be assigned by DHCPv6 and no matching
   PvD could be found, hosts MAY associate the assigned address with any
   Implicit PvD received on the same interface or to multiple Implicit
   PvDs received on the same interface.  This is intended to resolve
   backward-compatibility issues with rare deployments choosing to
   assign addresses with DHCPv6 while not sending any matching PIO.
   Implementations are suggested to flag or log such scenarios as errors
   to help detect misconfigurations.

3.4.2.  DHCPv4 Configuration Association

   Associating DHCPv4 [RFC2131] configuration elements with Explicit
   PvDs allows hosts to treat a set of IPv4 and IPv6 configurations as a
   single PvD with shared properties.  For example, consider a router
   that provides two different uplinks.  One could be a broadband
   network that has data rate and streaming properties described in PvD
   Additional Information and that provides both IPv4 and IPv6 network
   access.  The other could be a cellular network that provides only
   IPv6 network access and uses NAT64 [RFC6146].  The broadband network
   can be represented by an Explicit PvD that points to the Additional
   Information and also marks association with DHCPv4 information.  The
   cellular network can be represented by a different Explicit PvD that
   is not associated with DHCPv4.

   When a PvD-aware host retrieves configuration elements from DHCPv4,
   the information is associated either with a single Explicit PvD on
   that interface or else with all Implicit PvDs on the same interface.

   An Explicit PvD indicates its association with DHCPv4 information by
   setting the L-flag in the PvD Option.  If there is exactly one
   Explicit PvD that sets this flag, hosts MUST associate the DHCPv4
   information with that PvD.  Multiple Explicit PvDs on the same
   interface marking this flag is a misconfiguration, and hosts SHOULD
associate the DHCPv4 information with any Explicit PvD in this

   If no single Explicit PvD claims association with DHCPv4, the
   configuration elements coming from DHCPv4 MUST be associated with all
   Implicit PvDs identified by the interface on which the DHCPv4
   transaction happened.  This maintains existing host behavior.

3.4.3.  Connection Sharing by the Host

   The situation in which a host shares connectivity from an upstream
   interface (e.g., cellular) to a downstream interface (e.g., Wi-Fi) is
   known as 'tethering'.  Techniques such as ND Proxy [RFC4389], 64share
   [RFC7278], or prefix delegation (e.g., using DHCPv6-PD [RFC8415]) may
   be used for that purpose.

   Whenever the RAs received from the upstream interface contain a PvD
   Option, hosts that are sharing connectivity SHOULD include a PvD
   Option within the RAs sent downstream with:

   *  The same PvD ID FQDN

   *  The same H-flag, Delay, and Sequence Number values

   *  The L-flag set whenever the host is sharing IPv4 connectivity
      received from the same upstream interface

   *  The bits in the Reserved field set to 0

   The values of the R-flag, Router Advertisement message header, and
   Options field depend on whether or not the connectivity should be
   shared only with PvD-aware hosts (see Section 3.2).  In particular,
   all options received within the upstream PvD Option and included in
   the downstream RA SHOULD be included in the downstream PvD Option.

3.4.4.  Usage of DNS Servers

   PvD-aware hosts can be provisioned with recursive DNS servers via RA
   options passed within an Explicit PvD, via RA options associated with
   an Implicit PvD, via DHCPv6 or DHCPv4, or from some other
   provisioning mechanism that creates an Explicit PvD (such as a VPN).
   In all of these cases, the recursive DNS server addresses SHOULD be
   associated with the corresponding PvD.  Specifically, queries sent to
   a configured recursive DNS server SHOULD be sent from a local IP
   address that was provisioned for the PvD via RA or DHCP.  Answers
   received from the DNS server SHOULD only be used on the same PvD.

   PvD-aware applications will be able to select which PvD(s) to use for
   DNS resolution and connections, which allows them to effectively use
   multiple Explicit PvDs.  In order to support non-PvD-aware
   applications, however, PvD-aware hosts SHOULD ensure that non-PvD-
   aware name resolution APIs like "getaddrinfo" only use resolvers from
   a single PvD for a given query.  Handling DNS across PvDs is
   discussed in Section 5.2.1 of [RFC7556], and PvD APIs are discussed
   in Section 6 of [RFC7556].

   Maintaining the correct usage of DNS within PvDs avoids various
   practical errors such as:

   *  A PvD associated with a VPN or otherwise private network may
      provide DNS answers that contain addresses inaccessible over
      another PvD.  This includes the DNS queries to retrieve PvD
      Additional Information, which could otherwise send identifying
      information to the recursive DNS system (see Section 4.1).

   *  A PvD that uses a NAT64 [RFC6146] and DNS64 [RFC6147] will
      synthesize IPv6 addresses in DNS answers that are not globally
      routable and would be invalid on other PvDs.  Conversely, an IPv4
      address resolved via DNS on another PvD cannot be directly used on
      a NAT64 network.

4.  Provisioning Domain Additional Information

   Additional information about the network characteristics can be
   retrieved based on the PvD ID.  This set of information is called PvD
   Additional Information and is encoded as a JSON object [RFC8259].
   This JSON object is restricted to the Internet JSON (I-JSON) profile,
   as defined in [RFC7493].

   The purpose of this JSON object is to provide Additional Information
   to applications on a client host about the connectivity that is
   provided using a given interface and source address.  It typically
   includes data that would be considered too large, or not critical
   enough, to be provided within an RA option.  The information
   contained in this object MAY be used by the operating system, network
   libraries, applications, or users in order to decide which set of
   PvDs should be used for which connection, as described in
   Section 3.4.

   The Additional Information related to a PvD is specifically intended
   to be optional and is targeted at optimizing or informing the
   behavior of user-facing hosts.  This information can be extended to
   provide hints for host system behavior (such as captive portal or
   walled-garden PvD detection) or application behavior (describing
   application-specific services offered on a given PvD).  This content
   may not be appropriate for light-weight IoT devices.  IoT devices
   might need only a subset of the information and would in some cases
   prefer a smaller representation like Concise Binary Object
   Representation (CBOR) [RFC7049].  Delivering a reduced version of the
   PvD Additional Information designed for such devices is not defined
   in this document.

4.1.  Retrieving the PvD Additional Information

   When the H-flag of the PvD Option is set, hosts MAY attempt to
   retrieve the PvD Additional Information associated with a given PvD
   by performing an HTTP-over-TLS [RFC2818] GET query to "https://<PvD-
   ID>/.well-known/pvd".  Inversely, hosts MUST NOT do so whenever the
   H-flag is not set.

   Recommendations for how to use TLS securely can be found in

   When a host retrieves the PvD Additional Information, it MUST verify
   that the TLS server certificate is valid for the performed request,
   specifically, that a DNS-ID [RFC6125] on the certificate is equal to
   the PvD ID expressed as an FQDN.  This validation indicates that the
   owner of the FQDN authorizes its use with the prefix advertised by
   the router.  If this validation fails, hosts MUST close the
   connection and treat the PvD as if it has no Additional Information.

   HTTP requests and responses for PvD Additional Information use the
   "application/pvd+json" media type (see Section 8.5).  Clients SHOULD
   include this media type as an Accept header field in their GET
   requests, and servers MUST mark this media type as their Content-Type
   header field in responses.

   Note that the DNS name resolution of the PvD ID, any connections made
   for certificate validation (such as Online Certificate Status
   Protocol (OCSP) [RFC6960]), and the HTTP request itself MUST be
   performed using the considered PvD.  In other words, the name
   resolution, PKI checks, source address selection, as well as the
   next-hop router selection MUST be performed while exclusively using
   the set of configuration information attached with the PvD, as
   defined in Section 3.4.  In some cases, it may therefore be necessary
   to wait for an address to be available for use (e.g., once the
   Duplicate Address Detection or DHCPv6 processes are complete) before
   initiating the HTTP-over-TLS query.  In order to address privacy
   concerns around linkability of the PvD HTTP connection with future
   user-initiated connections, if the host has a temporary address per
   [RFC4941] in this PvD, then it SHOULD use a temporary address to
   fetch the PvD Additional Information and MAY deprecate the used
   temporary address and generate a new temporary address afterward.

   If the HTTP status of the answer is greater than or equal to 400, the
   host MUST close its connection and consider that there is no PvD
   Additional Information.  If the HTTP status of the answer is between
   300 and 399, inclusive, it MUST follow the redirection(s).  If the
   HTTP status of the answer is between 200 and 299, inclusive, the
   response is expected to be a single JSON object.

   After retrieval of the PvD Additional Information, hosts MUST
   remember the last Sequence Number value received in an RA including
   the same PvD ID.  Whenever a new RA for the same PvD is received with
   a different Sequence Number value, or whenever the expiry date for
   the additional information is reached, hosts MUST deprecate the
   Additional Information and stop using it.

   Hosts retrieving a new PvD Additional Information object MUST check
   for the presence and validity of the mandatory fields specified in
   Section 4.3.  A retrieved object including an expiration time that is
   already past or missing a mandatory element MUST be ignored.

   In order to avoid synchronized queries toward the server hosting the
   PvD Additional Information when an object expires, object updates are
   delayed by a randomized backoff time.

   *  When a host performs a JSON object update after it detected a
      change in the PvD Option Sequence Number, it MUST add a delay
      before sending the query.  The target time for the delay is
      calculated as a random time between zero and 2^((10 + Delay))
      milliseconds, where 'Delay' corresponds to the 4-bit unsigned
      integer in the last received PvD Option.

   *  When a host last retrieved a JSON object at time A that includes
      an expiry time B using the "expires" key, and the host is
      configured to keep the PvD Additional Information up to date, it
      MUST add some randomness into its calculation of the time to fetch
      the update.  The target time for fetching the updated object is
      calculated as a uniformly random time in the interval [(B-A)/2,B].

   In the example in Figure 2, the Delay field value is 1; this means
   that the host calculates its delay by choosing a uniformly random
   time between 0 and 2^((10 + 1)) milliseconds, i.e., between 0 and
   2048 milliseconds.

   Since the Delay value is directly within the PvD Option rather than
   the object itself, an operator may perform a push-based update by
   incrementing the Sequence Number value while changing the Delay value
   depending on the criticality of the update and the capacity of its
   PvD Additional Information servers.

   In addition to adding a random delay when fetching Additional
   Information, hosts MUST enforce a minimum time between requesting
   Additional Information for a given PvD on the same network.  This
   minimum time is RECOMMENDED to be 10 seconds, in order to avoid hosts
   causing a denial-of-service on the PvD server.  Hosts also MUST limit
   the number of requests that are made to different PvD Additional
   Information servers on the same network within a short period of
   time.  A RECOMMENDED value is to issue no more than five PvD
   Additional Information requests in total on a given network within 10
   seconds.  For more discussion, see Section 6.

   The PvD Additional Information object includes a set of IPv6 prefixes
   (under the key "prefixes") that MUST be checked against all the
   Prefix Information Options advertised in the RA.  If any of the
   prefixes included in any associated PIO is not covered by at least
   one of the listed prefixes, the PvD Additional Information MUST be
   considered to be a misconfiguration and MUST NOT be used by the host.
   See Section 4.4 for more discussion on handling such

   If the request for PvD Additional Information fails due to a TLS
   certificate validation error, an HTTP error, or because the retrieved
   file does not contain valid PvD JSON, hosts MUST close any connection
   used to fetch the PvD Additional Information and MUST NOT request the
   information for that PvD ID again for the duration of the local
   network attachment.  If a host detects 10 or more such failures to
   fetch PvD Additional Information, the local network is assumed to be
   misconfigured or under attack and the host MUST NOT make any further
   requests for any PvD Additional Information, belonging to any PvD ID,
   for the duration of the local network attachment.  For more
   discussion, see Section 6.

4.2.  Operational Consideration to Providing the PvD Additional

   Whenever the H-flag is set in the PvD Option, a valid PvD Additional
   Information object MUST be made available to all hosts receiving the
   RA by the network operator.  In particular, when a captive portal is
   present, hosts MUST still be allowed to perform DNS, certificate
   validation, and HTTP-over-TLS operations related to the retrieval of
   the object, even before logging into the captive portal.

   Routers SHOULD increment the PvD Option Sequence Number by one
   whenever a new PvD Additional Information object is available and
   should be retrieved by hosts.  If the value exceeds what can be
   stored in the Sequence Number field, it MUST wrap back to zero.

   The server providing the JSON files SHOULD also check whether the
   client address is contained by the prefixes listed in the Additional
   Information and SHOULD return a 403 response code if there is no

4.3.  PvD Additional Information Format

   The PvD Additional Information is a JSON object.

   The following table presents the mandatory keys, which MUST be
   included in the object:

    | JSON key   | Description   | Type      | Example                |
    | identifier | PvD ID FQDN   | String    | "pvd.example.com."     |
    | expires    | Date after    | [RFC3339] | "2020-05-23T06:00:00Z" |
    |            | which this    | Date      |                        |
    |            | object is no  |           |                        |
    |            | longer valid  |           |                        |
    | prefixes   | Array of IPv6 | Array of  | ["2001:db8:1::/48",    |
    |            | prefixes      | strings   | "2001:db8:4::/48"]     |
    |            | valid for     |           |                        |
    |            | this PvD      |           |                        |

                                  Table 1

   A retrieved object that does not include all three of these keys at
   the root of the JSON object MUST be ignored.  All three keys need to
   be validated; otherwise, the object MUST be ignored.  The value
   stored for "identifier" MUST be matched against the PvD ID FQDN
   presented in the PvD Option using the comparison mechanism described
   in Section 3.4.  The value stored for "expires" MUST be a valid date
   in the future.  If the PIO of the received RA is not covered by at
   least one of the "prefixes" key, the retrieved object SHOULD be

   The following table presents some optional keys that MAY be included
   in the object.

   | JSON key   | Description          | Type     | Example            |
   | dnsZones   | DNS zones searchable | Array    | ["example.com",    |
   |            | and accessible       | of       | "sub.example.com"] |
   |            |                      | strings  |                    |
   | noInternet | No Internet; set to  | Boolean  | true               |
   |            | "true" when the PvD  |          |                    |
   |            | is restricted        |          |                    |

                                  Table 2

   It is worth noting that the JSON format allows for extensions.
   Whenever an unknown key is encountered, it MUST be ignored along with
   its associated elements.

   Private-use or experimental keys MAY be used in the JSON dictionary.
   In order to avoid such keys colliding with the keys registered by
   IANA, implementers or vendors defining private-use or experimental
   keys MUST create sub-dictionaries.  If a set of PvD Additional
   Information keys are defined by an organization that has a formal URN
   namespace [IANA-URN], the URN namespace SHOULD be used as the top-
   level JSON key for the sub-dictionary.  For other private uses, the
   sub-dictionary key SHOULD follow the format of "vendor-*", where the
   "*" is replaced by the implementer's or vendor's identifier.  For
   example, keys specific to the FooBar organization could use "vendor-
   foobar".  If a host receives a sub-dictionary with an unknown key,
   the host MUST ignore the contents of the sub-dictionary.

4.3.1.  Example

   The following two examples show how the JSON keys defined in this
   document can be used:

     "identifier": "cafe.example.com.",
     "expires": "2020-05-23T06:00:00Z",
     "prefixes": ["2001:db8:1::/48", "2001:db8:4::/48"],

     "identifier": "company.foo.example.com.",
     "expires": "2020-05-23T06:00:00Z",
     "prefixes": ["2001:db8:1::/48", "2001:db8:4::/48"],
           "private-key": "private-value",

4.4.  Detecting Misconfiguration and Misuse

   Hosts MUST validate the TLS server certificate when retrieving PvD
   Additional Information, as detailed in Section 4.1.

   Hosts MUST verify that all prefixes in all the RA PIOs are covered by
   a prefix from the PvD Additional Information.  An adversarial router
   attempting to spoof the definition of an Explicit PvD, without the
   ability to modify the PvD Additional Information, would need to
   perform IPv6-to-IPv6 Network Prefix Translation (NPTv6) [RFC6296] in
   order to circumvent this check.  Thus, this check cannot prevent all
   spoofing, but it can detect misconfiguration or mismatched routers
   that are not adding a NAT.

   If NPTv6 is being added in order to spoof PvD ownership, the HTTPS
   server for Additional Information can detect this misconfiguration.
   The HTTPS server SHOULD validate the source addresses of incoming
   connections (see Section 4.1).  This check gives reasonable assurance
   that NPTv6 was not used and restricts the information to the valid
   network users.If the PvD does not provision IPv4 (it does not include
   the L-flag in the RA), the server cannot validate the source
   addresses of connections using IPv4.  Thus, the PvD ID FQDN for such
   PvDs SHOULD NOT have a DNS A record.

5.  Operational Considerations

   This section describes some example use cases of PvDs.  For the sake
   of simplicity, the RA messages will not be described in the usual
   ASCII art but rather in an indented list.  Values in the PvD Option
   header that are not included in the example are assumed to be zero or
   false (such as the H-flag, Sequence Number, and Delay fields).

5.1.  Exposing Extra RA Options to PvD-Aware Hosts

   In this example, there is one RA message sent by the router.  This
   message contains some options applicable to all hosts on the network
   and also a PvD Option that also contains other options only visible
   to PvD-aware hosts.

   *  RA Header: router lifetime = 6000

   *  Prefix Information Option: length = 4, prefix = 2001:db8:cafe::/64

   *  PvD Option header: length = 3 + 5 + 4, PvD ID FQDN = example.org.,
      R-flag = 0 (actual length of the header with padding 24 bytes = 3
      * 8 bytes)

      -  Recursive DNS Server: length = 5, addresses =
         [2001:db8:cafe::53, 2001:db8:f00d::53]

      -  Prefix Information Option: length = 4, prefix =

   Note that a PvD-aware host will receive two different prefixes,
   "2001:db8:cafe::/64" and "2001:db8:f00d::/64", both associated with
   the same PvD (identified by "example.org.").  A non-PvD-aware host
   will only receive one prefix, "2001:db8:cafe::/64".

5.2.  Different RAs for PvD-Aware and Non-PvD-Aware Hosts

   It is expected that for some years, networks will have a mixed
   environment of PvD-aware hosts and non-PvD-aware hosts.  If there is
   a need to give specific information to PvD-aware hosts only, then it
   is RECOMMENDED to send two RA messages, one for each class of hosts.
   This approach allows for two distinct sets of configuration
   information to be sent in a way that will not disrupt non-PvD-aware
   hosts.  It also lowers the risk that a single RA message will
   approach its MTU limit due to duplicated information.

   If two RA messages are sent for this reason, they MUST be sent from
   two different link-local source addresses (Section 3.2).  For
   example, here is the RA sent for non-PvD-aware hosts:

   *  RA Header: router lifetime = 6000 (non-PvD-aware hosts will use
      this router as a default router)

   *  Prefix Information Option: length = 4, prefix = 2001:db8:cafe::/64

   *  Recursive DNS Server Option: length = 3, addresses =

   *  PvD Option header: length = 3 + 2, PvD ID FQDN = foo.example.org.,
      R-flag = 1 (actual length of the header 24 bytes = 3 * 8 bytes)

      -  RA Header: router lifetime = 0 (PvD-aware hosts will not use
         this router as a default router), implicit length = 2

   And here is the RA sent for PvD-aware hosts:

   *  RA Header: router lifetime = 0 (non-PvD-aware hosts will not use
      this router as a default router)

   *  PvD Option header: length = 3 + 2 + 4 + 3, PvD ID FQDN =
      bar.example.org., R-flag = 1 (actual length of the header 24 bytes
      = 3 * 8 bytes)

      -  RA Header: router lifetime = 1600 (PvD-aware hosts will use
         this router as a default router), implicit length = 2

      -  Prefix Information Option: length = 4, prefix =

      -  Recursive DNS Server Option: length = 3, addresses =

   In the above example, non-PvD-aware hosts will only use the first
   listed RA sent by their default router and use the
   "2001:db8:cafe::/64" prefix.  PvD-aware hosts will autonomously
   configure addresses from both PIOs but will only use the source
   address in "2001:db8:f00d::/64" to communicate past the first-hop
   router since only the router sending the second RA will be used as
   the default router; similarly, they will use the DNS server
   "2001:db8:f00d::53" when communicating from this address.

5.3.  Enabling Multihoming for PvD-Aware Hosts

   In this example, the goal is to have one prefix from one RA be usable
   by both non-PvD-aware and PvD-aware hosts and to have another prefix
   usable only by PvD-aware hosts.  This allows PvD-aware hosts to be
   able to effectively multihome on the network.

   The first RA is usable by all hosts.  The only difference for PvD-
   aware hosts is that they can explicitly identify the PvD ID
   associated with the RA.  PvD-aware hosts will also use this prefix to
   communicate with non-PvD-aware hosts on the same network.

   *  RA Header: router lifetime = 6000 (non-PvD-aware hosts will use
      this router as a default router)

   *  Prefix Information Option: length = 4, prefix = 2001:db8:cafe::/64

   *  Recursive DNS Server Option: length = 3, addresses =

   *  PvD Option header: length = 3, PvD ID FQDN = foo.example.org.,
      R-flag = 0 (actual length of the header 24 bytes = 3 * 8 bytes)

   The second RA contains a prefix usable only by PvD-aware hosts.  Non-
   PvD-aware hosts will ignore this RA; hence, only the PvD-aware hosts
   will be multihomed.

   *  RA Header: router lifetime = 0 (non-PvD-aware hosts will not use
      this router as a default router)

   *  PvD Option header: length = 3 + 2 + 4 + 3, PvD ID FQDN =
      bar.example.org., R-flag = 1 (actual length of the header 24 bytes
      = 3 * 8 bytes)

      -  RA Header: router lifetime = 1600 (PvD-aware hosts will use
         this router as a default router), implicit length = 2

      -  Prefix Information Option: length = 4, prefix =

      -  Recursive DNS Server Option: length = 3, addresses =

   Note: the above examples assume that the router has received its PvD
   IDs from upstream routers or via some other configuration mechanism.
   Another document could define ways for the router to generate its own
   PvD IDs to allow the above scenario in the absence of PvD ID

5.4.  Providing Additional Information to PvD-Aware Hosts

   In this example, the router indicates that it provides Additional
   Information using the H-flag.  The Sequence Number on the PvD Option
   is set to 7 in this example.

   *  RA Header: router lifetime = 6000

   *  Prefix Information Option: length = 4, prefix = 2001:db8:cafe::/64

   *  Recursive DNS Server Option: length = 3, addresses =

   *  PvD Option header: length = 3, PvD ID FQDN = cafe.example.com.,
      Sequence Number = 7, R-flag = 0, H-flag = 1 (actual length of the
      header with padding 24 bytes = 3 * 8 bytes)

   A PvD-aware host will fetch <https://cafe.example.com/.well-known/
   pvd> to get the additional information.  The following example shows
   a GET request that the host sends, in HTTP/2 syntax [RFC7540]:

   :method = GET
   :scheme = https
   :authority = cafe.example.com
   :path = /.well-known/pvd
   accept = application/pvd+json

   The HTTP server will respond with the JSON Additional Information:

   :status = 200
   content-type = application/pvd+json
   content-length = 116

     "identifier": "cafe.example.com.",
     "expires": "2020-05-23T06:00:00Z",
     "prefixes": ["2001:db8:cafe::/48"],

   At this point, the host has the PvD Additional Information and knows
   the expiry time.  When either the expiry time passes or a new
   Sequence Number is provided in an RA, the host will re-fetch the
   Additional Information.

   For example, if the router sends a new RA with the Sequence Number
   set to 8, the host will re-fetch the Additional Information:

   *  PvD Option header: length = 3 + 5 + 4 , PvD ID FQDN =
      cafe.example.com., Sequence Number = 8, R-flag = 0, H-flag = 1
      (actual length of the header with padding 24 bytes = 3 * 8 bytes)

   However, if the router sends a new RA, but the Sequence Number has
   not changed, the host would not re-fetch the Additional Information
   (until and unless the expiry time of the Additional Information has

6.  Security Considerations

   Since the PvD Option can contain an RA header and other RA options,
   any security considerations that apply for specific RA options
   continue to apply when used within a PvD Option.

   Although some solutions such as IPsec or SEcure Neighbor Discovery
   (SeND) [RFC3971] can be used in order to secure the IPv6 Neighbor
   Discovery Protocol, in practice, actual deployments largely rely on
   link-layer or physical-layer security mechanisms (e.g., 802.1x
   [IEEE8021X]) in conjunction with RA-Guard [RFC6105].

   If multiple RAs are sent for a single PvD to avoid fragmentation,
   dropping packets can lead to processing only part of a PvD Option,
   which could lead to hosts receiving only part of the contained
   options.  As discussed in Section 3.2, routers MUST include the PvD
   Option in all fragments generated.

   This specification does not improve the Neighbor Discovery Protocol
   security model but simply validates that the owner of the PvD FQDN
   authorizes its use with the prefix advertised by the router.  In
   combination with implicit trust in the local router (if present),
   this gives the host some level of assurance that the PvD is
   authorized for use in this environment.  However, when the local
   router cannot be trusted, no such guarantee is available.

   It must be noted that Section 4.4 of this document only provides
   reasonable assurance against misconfiguration but does not prevent a
   hostile network access provider from advertising incorrect
   information that could lead applications or hosts to select a hostile
   PvD.  However, a host that correctly implements the multiple PvD
   architecture [RFC7556] using the mechanism described in this document
   will be less susceptible to some attacks than a host that does not by
   being able to check for the various misconfigurations or
   inconsistencies described in this document.

   Since expiration times provided in PvD Additional Information use
   absolute time, these values can be skewed due to clock skew or for
   hosts without an accurate time base.  Such time values MUST NOT be
   used for security-sensitive functionality or decisions.

   An attacker generating RAs on a local network can use the H-flag and
   the PvD ID to cause hosts on the network to make requests for PvD
   Additional Information from servers.  This can become a denial-of-
   service attack, in which an attacker can amplify its attack by
   triggering TLS connections to arbitrary servers in response to
   sending UDP packets containing RA messages.  To mitigate this attack,
   hosts MUST:

   *  limit the rate at which they fetch a particular PvD's Additional

   *  limit the rate at which they fetch any PvD Additional Information
      on a given local network;

   *  stop making requests for a PvD ID that does not respond with valid
      JSON; and

   *  stop making requests for all PvD IDs once a certain number of
      failures is reached on a particular network.

   Details are provided in Section 4.1.  This attack can be targeted at
   generic web servers, in which case the host behavior of stopping
   requesting for any server that doesn't behave like a PvD Additional
   Information server is critical.  Limiting requests for a specific PvD
   ID might not be sufficient if the attacker changes the PvD ID values
   quickly, so hosts also need to stop requesting if they detect
   consistent failure when on a network that is under attack.  For cases
   in which an attacker is pointing hosts at a valid PvD Additional
   Information server (but one that is not actually associated with the
   local network), the server SHOULD reject any requests that do not
   originate from the expected IPv6 prefix as described in Section 4.2.

7.  Privacy Considerations

   Retrieval of the PvD Additional Information over HTTPS requires early
   communications between the connecting host and a server that may be
   located further than the first-hop router.  Although this server is
   likely to be located within the same administrative domain as the
   default router, this property can't be ensured.  To minimize the
   leakage of identity information while retrieving the PvD Additional
   Information, hosts SHOULD make use of an IPv6 temporary address and
   SHOULD NOT include any privacy-sensitive data, such as a User-Agent
   header field or an HTTP cookie.

   Hosts might not always fetch PvD Additional Information, depending on
   whether or not they expect to use the information.  However, if a
   host allows requesting Additional Information for certain PvD IDs, an
   attacker could send various PvD IDs in RAs to detect which PvD IDs
   are allowed by the client.  To avoid this, hosts SHOULD either fetch
   Additional Information for all eligible PvD IDs on a given local
   network or fetch the information for none of them.

   From a user privacy perspective, retrieving the PvD Additional
   Information is not different from establishing a first connection to
   a remote server or even performing a single DNS lookup.  For example,
   most operating systems already perform early queries to static web
   sites, such as <http://captive.example.com/hotspot-detect.html>, in
   order to detect the presence of a captive portal.

   The DNS queries associated with the PvD Additional Information MUST
   use the DNS servers indicated by the associated PvD, as described in
   Section 4.1.  This ensures the name of the PvD Additional Information
   server is not unintentionally sent on another network, thus leaking
   identifying information about the networks with which the client is

   There may be some cases where hosts, for privacy reasons, should
   refrain from accessing servers that are located outside a certain
   network boundary.  In practice, this could be implemented as an
   allowed list of 'trusted' FQDNs and/or IP prefixes that the host is
   allowed to communicate with.  In such scenarios, the host SHOULD
   check that the provided PvD ID, as well as the IP address that it
   resolves into, are part of the allowed list.

   Network operators SHOULD restrict access to PvD Additional
   Information to only expose it to hosts that are connected to the
   local network, especially if the Additional Information would provide
   information about local network configuration to attackers.  This can
   be implemented by allowing access from the addresses and prefixes
   that the router provides for the PvD, which will match the prefixes
   contained in the PvD Additional Information.  This technique is
   described in Section 4.2.

8.  IANA Considerations

8.1.  Change to IPv6 Neighbor Discovery Option Formats Registry

   IANA has removed the 'reclaimable' tag for value 21 for the PvD
   Option in the "IPv6 Neighbor Discovery Option Formats" registry.

8.2.  New Entry in the Well-Known URIs Registry

   IANA has added a new entry in the "Well-Known URIs" registry
   [RFC8615] with the following information:

   URI suffix: pvd

   Change controller: IETF

   Specification document: RFC 8801

   Status: permanent

   Related information: N/A

8.3.  New Additional Information PvD Keys Registry

   IANA has created and will maintain a new registry called "Additional
   Information PvD Keys", which reserves JSON keys for use in PvD
   Additional Information.  The initial contents of this registry are
   given in Section 4.3 (both the table of mandatory keys and the table
   of optional keys).

   The status of a key as mandatory or optional is intentionally not
   denoted in the table to allow for flexibility in future use cases.
   Any new assignments of keys will be considered as optional for the
   purpose of the mechanism described in this document.

   New assignments in the "Additional Information PvD Keys" registry
   will be administered by IANA through Expert Review [RFC8126].
   Experts are requested to ensure that defined keys do not overlap in
   names or semantics and that they represent non-vendor-specific use
   cases.  Vendor-specific keys SHOULD use sub-dictionaries, as
   described in Section 4.3.

   IANA has placed the "Additional Information PvD Keys" registry within
   a new registry entitled "Provisioning Domains (PvDs)".

8.4.  New PvD Option Flags Registry

   IANA has also created and will maintain a new registry entitled "PvD
   Option Flags".  This new registry reserves bit positions from 0 to 11
   to be used in the PvD Option bitmask.  This document assigns bit
   positions 0, 1, and 2 as shown in the table below.  Future
   assignments require Standards Action [RFC8126].

                     | Bit  | Name       | Reference |
                     | 0    | H-flag     | RFC 8801  |
                     | 1    | L-flag     | RFC 8801  |
                     | 2    | R-flag     | RFC 8801  |
                     | 3-11 | Unassigned |           |

                                  Table 3

   Since these flags apply to an IPv6 Router Advertisement Option, IANA
   has placed this registry under the existing "Internet Control Message
   Protocol version 6 (ICMPv6) Parameters" registry and provided a link
   on the new "Provisioning Domains (PvDs)" registry.

8.5.  PvD JSON Media Type Registration

   This document registers the media type for PvD JSON text,

   Type name:  application

   Subtype name:  pvd+json

   Required parameters:  N/A

   Optional parameters:  N/A

   Encoding considerations:  Encoding considerations are identical to
      those specified for the "application/json" media type.

   Security considerations:  See Section 6 of RFC 8801.

   Interoperability considerations:  This document specifies the format
      of conforming messages and the interpretation thereof.

   Published specification:  RFC 8801

   Applications that use this media type:  This media type is intended
      to be used by networks advertising additional Provisioning Domain
      information and clients looking up such information.

   Fragment identifier considerations:  N/A

   Additional information:  N/A

   Person & email address to contact for further information:  See
      Authors' Addresses section

   Intended usage:  COMMON

   Restrictions on usage:  N/A

   Author:  IETF

   Change controller:  IETF

9.  References

9.1.  Normative References

   [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,

   [RFC2818]  Rescorla, E., "HTTP Over TLS", RFC 2818,
              DOI 10.17487/RFC2818, May 2000,

   [RFC3339]  Klyne, G. and C. Newman, "Date and Time on the Internet:
              Timestamps", RFC 3339, DOI 10.17487/RFC3339, July 2002,

   [RFC4191]  Draves, R. and D. Thaler, "Default Router Preferences and
              More-Specific Routes", RFC 4191, DOI 10.17487/RFC4191,
              November 2005, <https://www.rfc-editor.org/info/rfc4191>.

   [RFC4343]  Eastlake 3rd, D., "Domain Name System (DNS) Case
              Insensitivity Clarification", RFC 4343,
              DOI 10.17487/RFC4343, January 2006,

   [RFC4861]  Narten, T., Nordmark, E., Simpson, W., and H. Soliman,
              "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861,
              DOI 10.17487/RFC4861, September 2007,

   [RFC4941]  Narten, T., Draves, R., and S. Krishnan, "Privacy
              Extensions for Stateless Address Autoconfiguration in
              IPv6", RFC 4941, DOI 10.17487/RFC4941, September 2007,

   [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,

   [RFC6980]  Gont, F., "Security Implications of IPv6 Fragmentation
              with IPv6 Neighbor Discovery", RFC 6980,
              DOI 10.17487/RFC6980, August 2013,

   [RFC7493]  Bray, T., Ed., "The I-JSON Message Format", RFC 7493,
              DOI 10.17487/RFC7493, March 2015,

   [RFC7525]  Sheffer, Y., Holz, R., and P. Saint-Andre,
              "Recommendations for Secure Use of Transport Layer
              Security (TLS) and Datagram Transport Layer Security
              (DTLS)", BCP 195, RFC 7525, DOI 10.17487/RFC7525, May
              2015, <https://www.rfc-editor.org/info/rfc7525>.

   [RFC7556]  Anipko, D., Ed., "Multiple Provisioning Domain
              Architecture", RFC 7556, DOI 10.17487/RFC7556, June 2015,

   [RFC8028]  Baker, F. and B. Carpenter, "First-Hop Router Selection by
              Hosts in a Multi-Prefix Network", RFC 8028,
              DOI 10.17487/RFC8028, November 2016,

   [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,

   [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>.

   [RFC8259]  Bray, T., Ed., "The JavaScript Object Notation (JSON) Data
              Interchange Format", STD 90, RFC 8259,
              DOI 10.17487/RFC8259, December 2017,

   [RFC8615]  Nottingham, M., "Well-Known Uniform Resource Identifiers
              (URIs)", RFC 8615, DOI 10.17487/RFC8615, May 2019,

9.2.  Informative References

   [IANA-URN] IANA, "Uniform Resource Names (URN) Namespaces",

              IEEE, "IEEE Standard for Local and Metropolitan Area
              Networks -- Port-Based Network Access Control", IEEE
              802.1X-2020, DOI 10.1109/IEEESTD.2020.9018454,

   [MPVD-API] Kline, E., "Multiple Provisioning Domains API
              Requirements", Work in Progress, Internet-Draft, draft-
              kline-mif-mpvd-api-reqs-00, 1 November 2015,

   [MPVD-DNS] Stenberg, M. and S. Barth, "Multiple Provisioning Domains
              using Domain Name System", Work in Progress, Internet-
              Draft, draft-stenberg-mif-mpvd-dns-00, 15 October 2015,

   [RFC2131]  Droms, R., "Dynamic Host Configuration Protocol",
              RFC 2131, DOI 10.17487/RFC2131, March 1997,

   [RFC3646]  Droms, R., Ed., "DNS Configuration options for Dynamic
              Host Configuration Protocol for IPv6 (DHCPv6)", RFC 3646,
              DOI 10.17487/RFC3646, December 2003,

   [RFC3971]  Arkko, J., Ed., Kempf, J., Zill, B., and P. Nikander,
              "SEcure Neighbor Discovery (SEND)", RFC 3971,
              DOI 10.17487/RFC3971, March 2005,

   [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery
              Proxies (ND Proxy)", RFC 4389, DOI 10.17487/RFC4389, April
              2006, <https://www.rfc-editor.org/info/rfc4389>.

   [RFC6105]  Levy-Abegnoli, E., Van de Velde, G., Popoviciu, C., and J.
              Mohacsi, "IPv6 Router Advertisement Guard", RFC 6105,
              DOI 10.17487/RFC6105, February 2011,

   [RFC6125]  Saint-Andre, P. and J. Hodges, "Representation and
              Verification of Domain-Based Application Service Identity
              within Internet Public Key Infrastructure Using X.509
              (PKIX) Certificates in the Context of Transport Layer
              Security (TLS)", RFC 6125, DOI 10.17487/RFC6125, March
              2011, <https://www.rfc-editor.org/info/rfc6125>.

   [RFC6146]  Bagnulo, M., Matthews, P., and I. van Beijnum, "Stateful
              NAT64: Network Address and Protocol Translation from IPv6
              Clients to IPv4 Servers", RFC 6146, DOI 10.17487/RFC6146,
              April 2011, <https://www.rfc-editor.org/info/rfc6146>.

   [RFC6147]  Bagnulo, M., Sullivan, A., Matthews, P., and I. van
              Beijnum, "DNS64: DNS Extensions for Network Address
              Translation from IPv6 Clients to IPv4 Servers", RFC 6147,
              DOI 10.17487/RFC6147, April 2011,

   [RFC6296]  Wasserman, M. and F. Baker, "IPv6-to-IPv6 Network Prefix
              Translation", RFC 6296, DOI 10.17487/RFC6296, June 2011,

   [RFC6960]  Santesson, S., Myers, M., Ankney, R., Malpani, A.,
              Galperin, S., and C. Adams, "X.509 Internet Public Key
              Infrastructure Online Certificate Status Protocol - OCSP",
              RFC 6960, DOI 10.17487/RFC6960, June 2013,

   [RFC7049]  Bormann, C. and P. Hoffman, "Concise Binary Object
              Representation (CBOR)", RFC 7049, DOI 10.17487/RFC7049,
              October 2013, <https://www.rfc-editor.org/info/rfc7049>.

   [RFC7278]  Byrne, C., Drown, D., and A. Vizdal, "Extending an IPv6
              /64 Prefix from a Third Generation Partnership Project
              (3GPP) Mobile Interface to a LAN Link", RFC 7278,
              DOI 10.17487/RFC7278, June 2014,

   [RFC7540]  Belshe, M., Peon, R., and M. Thomson, Ed., "Hypertext
              Transfer Protocol Version 2 (HTTP/2)", RFC 7540,
              DOI 10.17487/RFC7540, May 2015,

   [RFC8106]  Jeong, J., Park, S., Beloeil, L., and S. Madanapalli,
              "IPv6 Router Advertisement Options for DNS Configuration",
              RFC 8106, DOI 10.17487/RFC8106, March 2017,

   [RFC8415]  Mrugalski, T., Siodelski, M., Volz, B., Yourtchenko, A.,
              Richardson, M., Jiang, S., Lemon, T., and T. Winters,
              "Dynamic Host Configuration Protocol for IPv6 (DHCPv6)",
              RFC 8415, DOI 10.17487/RFC8415, November 2018,


   Many thanks to Markus Stenberg and Steven Barth for their earlier
   work on [MPVD-DNS], as well as to Basile Bruneau, who was author of
   an early draft version of this document.

   Thanks also to Marcus Keane, Mikael Abrahamsson, Ray Bellis, Zhen
   Cao, Tim Chown, Lorenzo Colitti, Michael Di Bartolomeo, Ian Farrer,
   Phillip Hallam-Baker, Bob Hinden, Tatuya Jinmei, Erik Kline, Ted
   Lemon, Paul Hoffman, Dave Thaler, Suresh Krishnan, Gorry Fairhurst,
   Jen Lenkova, Veronika McKillop, Mark Townsley, and James Woodyatt for
   useful and interesting discussions and reviews.

   Finally, special thanks to Thierry Danis for his valuable input and
   implementation efforts, Tom Jones for his integration effort into the
   NEAT project, and Rigil Salim for his implementation work.

Authors' Addresses

   Pierre Pfister
   11 Rue Camille Desmoulins
   92130 Issy-les-Moulineaux

   Email: ppfister@cisco.com

   Éric Vyncke
   De Kleetlaan, 6
   1831 Diegem

   Email: evyncke@cisco.com

   Tommy Pauly
   Apple Inc.
   One Apple Park Way
   Cupertino, California 95014
   United States of America

   Email: tpauly@apple.com

   David Schinazi
   Google LLC
   1600 Amphitheatre Parkway
     View, California 94043
   United States of America

   Email: dschinazi.ietf@gmail.com

   Wenqin Shao
   11 Rue Camille Desmoulins
   92130 Issy-les-Moulineaux

   Email: wenshao@cisco.com