path: root/draft-schanzen-r5n.xml

<?xml version='1.0' encoding='utf-8'?>
<!DOCTYPE rfc SYSTEM "rfc2629-xhtml.ent" [
<!ENTITY RFC2119 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2119.xml">
<!ENTITY RFC2782 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2782.xml">
<!ENTITY RFC3629 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3629.xml">
<!ENTITY RFC3686 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3686.xml">
<!ENTITY RFC3826 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3826.xml">
<!ENTITY RFC3912 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3912.xml">
<!ENTITY RFC3986 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3986.xml">
<!ENTITY RFC4648 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4648.xml">
<!ENTITY RFC5869 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5869.xml">
<!ENTITY RFC5890 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5890.xml">
<!ENTITY RFC5891 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5891.xml">
<!ENTITY RFC6781 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.6781.xml">
<!ENTITY RFC6895 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.6895.xml">
<!ENTITY RFC6940 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.6940.xml">
<!ENTITY RFC6979 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.6979.xml">
<!ENTITY RFC7748 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.7748.xml">
<!ENTITY RFC8032 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.8032.xml">
<!ENTITY RFC8126 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.8126.xml">
<!ENTITY RFC8174 PUBLIC '' "http://xml.resource.org/public/rfc/bibxml/reference.RFC.8174.xml">
]>
<?xml-stylesheet type='text/xsl' href='rfc2629.xslt' ?>
<?rfc strict="yes" ?>
<?rfc toc="yes" ?>
<?rfc symrefs="yes"?>
<?rfc sortrefs="yes" ?>
<?rfc compact="yes" ?>
<?rfc subcompact="no" ?>
<rfc xmlns:xi="http://www.w3.org/2001/XInclude" category="info" docName="draft-schanzen-r5n-00" ipr="trust200902" obsoletes="" updates="" submissionType="IETF" xml:lang="en" version="3">
 <!-- xml2rfc v2v3 conversion 2.26.0 -->
 <front>
  <title abbrev="The R5N Distributed Hash Table">
   The R5N Distributed Hash Table
  </title>
  <seriesInfo name="Internet-Draft" value="draft-schanzen-r5n-00"/>
  <author fullname="Martin Schanzenbach" initials="M." surname="Schanzenbach">
   <organization>GNUnet e.V.</organization>
   <address>
    <postal>
     <street>Boltzmannstrasse 3</street>
     <city>Garching</city>
     <code>85748</code>
     <country>DE</country>
    </postal>
    <email>schanzen@gnunet.org</email>
   </address>
  </author>
  <author fullname="Christian Grothoff" initials="C." surname="Grothoff">
   <organization>Berner Fachhochschule</organization>
   <address>
    <postal>
     <street>Hoeheweg 80</street>
     <city>Biel/Bienne</city>
     <code>2501</code>
     <country>CH</country>
    </postal>
    <email>grothoff@gnunet.org</email>
   </address>
  </author>
  <author fullname="Bernd Fix" initials="B." surname="Fix">
   <organization>GNUnet e.V.</organization>
   <address>
    <postal>
     <street>Boltzmannstrasse 3</street>
     <city>Garching</city>
     <code>85748</code>
     <country>DE</country>
    </postal>
    <email>fix@gnunet.org</email>
   </address>
  </author>

  <!-- Meta-data Declarations -->
  <area>General</area>
  <workgroup>Independent Stream</workgroup>
  <keyword>distributed hash tables</keyword>
  <abstract>
    <t>This document contains the R5N DHT technical specification.</t>
    <t>
      This document defines the normative wire format of resource records,
      resolution processes, cryptographic routines and security considerations for
      use by implementers.
    </t>
    <t>
      This specification was developed outside the IETF and does not have IETF
      consensus. It is published here to guide implementation of R5N and to
      ensure interoperability among implementations.
    </t>
  </abstract>
 </front>
 <middle>
   <section anchor="introduction" numbered="true" toc="default">
     <name>Introduction</name>
     <!-- FIXME: Here we should also cite and discuss RELOAD (https://datatracker.ietf.org/doc/html/rfc6940)
       and establish why we need this spec and are not a "Topology plugin"
       in RELOAD. The argumentation revolves around the trust model (openness) and
       security aspects (path signatures).
     -->
     <t>
       Distributed Hash Tables (DHTs) are a key data structure for the
       construction of completely decentralized applications.
       DHTs are important because they generally provide a robust and
       efficient means to distribute the storage and retrieval of
       key-value pairs.
     </t>
     <t>
       While <xref target="RFC6940" /> already provides a peer-to-peer (P2P)
       signaling protocol with extensible routing and topology mechanisms,
       it also relies on strict admission control through the use of either
       centralized enrollment servers or pre-shared keys.
       Modern decentralized applications require a more open system that
       enables ad-hoc participation and other means to prevent common attacks
       on P2P overlays.
     </t>
     <t>
       This document contains the technical specification
       of the R5N DHT <xref target="R5N" />, a secure DHT routing algorithm
       and data structure for decentralized applications.
       R5N is an open P2P overlay routing mechanism which supports ad-hoc
       participation and security properties including support for
       topologies in restricted-route environments and path signatures.
     </t>
     <t>
       This document defines the normative wire format of peer-to-peer
       messages, routing algorithms, cryptographic routines and security
       considerations for use by implementors.
     </t>
     <section numbered="true" toc="default">
       <name>Requirements Notation</name>
       <t>
         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 <xref target="RFC2119"/> <xref target="RFC8174"/> when, and only
         when, they appear in all capitals, as shown here.
       </t>
     </section>

   </section>
   <section anchor="architecture" numbered="true" toc="default">
     <name>Architecture</name>
     <t>
       R5N is an overlay network with a pluggable transport layer.
       The following figure shows the R5N architecture.
     </t>
       <figure>
         <artwork name="" type="" align="left" alt=""><![CDATA[
             |  +-----------------+  +-------+
Applications |  | GNU Name System |  | CADET |  ...
             |  +-----------------+  +-------+
-------------+------------------------------------ Overlay Interface
             |  ^
             |  |   +---------------+
             |  |   | Block Storage |
             |  |   +---------------+
             |  |    ^
R5N          |  v    v
             | +--------------------+    +---------+
             | | Message Processing |<-->| Routing |
             | +--------------------+    +---------+
             |  ^                          ^
             |  v                          v
-------------+------------------------------------ Underlay Interface
             | +--------+  +--------+
             | |GNUnet  |  |IP      |  ...
Connectivity | |Underlay|  |Underlay|
             | |Link    |  |Link    |
             | +--------+  +--------+
           ]]></artwork>
       </figure>
     <dl>
       <dt>Applications</dt>
       <dd>
         Applications are components which directly use the DHT overlay
         interfaces. Possible applications include the GNU Name System
         <xref target="I-D.draft-schanzen-gns" /> or the CADET transport system
         <xref target="cadet" />.
       </dd>
       <dt>Overlay Interface</dt>
       <dd>
         The Overlay Interface exposes the core operations of the DHT overlay
         to applications.
         This includes querying and retrieving data from the DHT.
       </dd>
       <dt>Block Storage</dt>
       <dd>
         The Block Storage component is used to persist and manage data
         by peers. It includes logic for quotas, caching stragegies and
         data validation.
       </dd>
       <dt>Message Processing</dt>
       <dd>
         The Message Processing component processes requests from and responses
         to applications as well as messages from the underlay network.
       </dd>
       <dt>Routing</dt>
       <dd>
         The Routing component includes the routing table as well as
         routing and peer selection logic. It facilitates the R5N routing
         algorithm with required data structures and algorithms.
       </dd>
       <dt>Underlay Interface</dt>
       <dd>
         The DHT Underlay Interface is an abstraction layer on top of the
         supported links of a peer. Peers may be linked by a variety of
         different transports, including "classical" protocols such as
         TCP, UDP and TLS or advanced protocols such as GNUnet, L2P or Tor.
       </dd>
     </dl>
   </section>
   <section anchor="overlay" numbered="true" toc="default">
     <name>Overlay</name>
     <t>
       In the DHT overlay, a peer is addressable by its Peer ID.
       The Peer ID is the 256-bit hash of the peer public key.
       The peer public key is the public key of the corresponding
       Ed25519<xref target="ed25519" /> peer private key.
     </t>
     <t>
       Any implementation of this specification MUST expose the two API
       procedures "GET" and "PUT".
     </t>
     <section>
       <name>The GET procedure</name>
       <t>
         The GET procedure is defined as follows:
       </t>
         <artwork name="" type="" align="left" alt=""><![CDATA[
GET(key[, options]) -> RESULTS as List
           or
GET(key[, options], callbackFunction)
         ]]></artwork>
       <t>
         The procedure takes two arguments. The first argument is the query
         key and is mandatory. The GET procedure may also allow the caller
         to specifiy query options.
       </t>
       <t>
         The procedure either returns a list of results or allows the caller
         to provide a callback function which is called for any result
         received from the DHT until the procedure is cancelled.
       </t>
     </section>
     <section>
       <name>The PUT procedure</name>
       <t>
         The PUT procedure is defined as follows:
       </t>
         <artwork name="" type="" align="left" alt=""><![CDATA[
PUT(key[, options], BLOCK)
         ]]></artwork>
       <t>
         The procedure takes three arguments. The first argument is the query
         key and is mandatory. The PUT procedure may also allow the caller
         to specifiy put options. The third argument is the payload data which
         is to be stored under the provided put key.
       </t>
     </section>
   </section>
   <section anchor="underlay" numbered="true" toc="default">
     <name>Underlay</name>
     <t>
       In the network underlay, a peer is addressable by traditional
       means out of scope of this document. For example, the peer may
       have a TCP/IP address, or a HTTPS endpoint.
       While the specific addressing options and mechanisms are out of scope
       for this document, it is necessary to define a universal addressing
       format in order to facilitate the distribution of connectivity
       information to other peers in the DHT overlay.
       This format is the "HELLO" message. A "HELLO" is a human-readable
       UTF-8  <xref target="RFC3629" /> string consisting of the peer
       public key and the HELLO URI <xref target="RFC3986" />.
     </t>
       <figure>
         <artwork name="" type="" align="left" alt=""><![CDATA[
hello-format := <peer-public-key> <hello-uri>
peer-public-key := [A-HJ-NP-Z1-9]+
         ]]></artwork>
     </figure>
     <t>
       For the string representation of the peer public key,
       the base-32 encoding "StringEncode" is used.
       However, instead of following <xref target="RFC4648"/> the
       character map is based on the optical character recognition friendly
       proposal of Crockford <xref target="CrockfordB32"/>.
       The only difference to Crockford is that the letter
       "U" decodes to the same base-32 value as the letter "V" (27).
     </t>
     <t>
       The "scheme" part of the HELLO URI defined the addressing scheme
       which is used. An example of an addressing scheme used throughout
       this document is "ip+tcp", which refers to a standard TCP/IP socket
       connection. The "hier"-part of the URI must provide a suitable
       address for the given addressing scheme.
       The following is a non-normative example of a HELLO containing three
       HELLO URIs:
     </t>
     <!-- FIXME peer id type | length | id payload | 0-terminated strings for addresses -->
       <figure>
         <artwork name="" type="" align="left" alt=""><![CDATA[
Y924NSHMMZ1N1SQCE5TXF93ED6S6JY311K0QT86G9WJC68F6XVZ0 \
        ip+tcp://1.2.3.4:6789 \
        gnunet+tcp://12.3.4.5/ \
        i2p+udp://1.2.4.5:424/ \
        tor+onionv3://rasdflkjasdfliasduf.onion/
         ]]></artwork>
     </figure>


     <!--
       1) The current API is always fire+forget, it doesn't allow for flow
       control. I think we need to add that, possibly for sending and receiving.

       IDK.

2) I'm not sure what to do with the crypto: mandate EdDSA or allow the
       underlay to do whatever public keys it likes.

       We need keys in the overlay. (Path signatures). Do they need to
       be the same keys???

3) I think we may want to mandate that the lower layer at least
authenticate the other peer (i.e. every UDP message could be in
cleartext, but would need to come with an EdDSA signature, alas 92 byte
overhead and a signature verification _required_).  Otherwise, I don't
see how we can offer even the most minimal protections against peer
       impersonation attacks. WDYT?

       Security considerations? Prerequisites?
     -->
     <t>
       It is expected that there are basic mechanisms available to
       manage peer connectivity and addressing.
       The required functionality are abstracted through the following
       procedures and events:
     </t>
     <dl>
       <dt>PEER_CONNECTED(phash,address)</dt>
       <dd>
         is a signal that allows the DHT to react to peers which connect.
         Such an event triggers, for example, updates in the
         routing table.
       </dd>
       <dt>PEER_DISCONNECTED(phash,address)</dt>
       <dd>
         is a signal that allows the DHT to react to peers which disconnect.
         Such an event triggers, for example, updates in the
         routing table.
       </dd>
       <dt>TRY_CONNECT(pid, address)</dt>
       <dd>
         A function which allows a peer to attempt the establishment of
         a connection to another peer using an address.
       </dd>
       <dt>HOLD(pash)</dt>
       <dd>
         A function which tells the underlay to keep a hold on the connection
         to another peer.
       </dd>
       <dt>DROP(pash)</dt>
       <dd>
         A function which tells the underlay to drop the connection to another
         peer.
       </dd>
       <dt>RECEIVE(source, message)</dt>
       <dd>
         A function or event that allows the peer to receive protocol
         messages as defined in this document from a connected peer.
       </dd>
       <dt>SEND(target, message)</dt>
       <dd>
         A function that allows a peer to send protocol messages as defined
         in this document to a connected peer. If call to SEND fails,
         the message has not been sent.
       </dd>
       <dt>NETWORK_SIZE_ESTIMATE(N)</dt>
       <dd>
         A function or event that provides estimates on the network size
         for use in the DHT routing algorithms.
       </dd>
       <dt>ADDRESS_ADD(pk, address)</dt>
       <dd>
         The underlay signals us that an address was added.
         This information is used, for example, to publish
         connectivity as part of the bootstrapping and overlay creation.
       </dd>
       <dt>ADDRESS_DELETE(pk, address)</dt>
       <dd>
         The underlay signals us that an address was removed.
         This information is used, for example, to publish
         connectivity as part of the bootstrapping and overlay creation.
       </dd>
       <dt>VERIFY(blob)</dt>
       <dd>
         Signature verification by underlay.
       </dd>
     </dl>
   </section>

   <section anchor="routing" numbered="true" toc="default">
     <name>Routing</name>
     <section anchor="peer_selection" numbered="true" toc="default">
       <name>Peer selection</name>
       <t>
         In order to select peers from the routing table which are suitable
         destinations for sending messages, R5N uses a hybrid approach:
         Given an estimated network size N, the peer selection for the
         first N hops is random. After the initial N hops, peer selection
         follows an XOR-based peer distance calculation.
       </t>
       <t>
         As the message traverses a random path through the network for the
         first N hops, it is essential that routing loops are avoided.
         In R5N, a bloomfilter is used as part of the routing metadata in
         messages. The bloomfilter is updates at each hop with the hops
         peer identity.
         For the next hop selection in both the random and the deterministic
         case, any peer which is in the bloomfilter for the respective message
         is not included in the peer selection process.
       </t>
       <!-- Fixme: We may want to propose our modified, optimized XOR metric here or reference Kademlia -->
       <t>
         R5N stores the information of all connected peers in a a set of lists
         similar to the k-buckets data structure of <xref target="Kademlia"/>.
         The index which determines in which of the k lists to add a given peer
         is calculated using the FIND-BUCKET procedure (see <xref target="find-bucket"/>.
       </t>
       <t>
         The buckets serve implicitly as a routing table for messages:
         In order to select a peer for a given message key and bloomfilter,
         the <tt>PEER-SELECT</tt> is used (see <xref target="peer-select"/>.
       </t>
       <figure anchor="peer-select">
         <artwork name="" type="" align="left" alt=""><![CDATA[
PEER-SELECT(key, bloomfilter)
  peers := <Select all known peers NOT in message bloomfilter>
  IF hops >= N
    dist := MAX_VALUE
    FOR EACH p IN peers
      IF XOR(p, key) < dist
        dist := XOR(p, key)
        target := p
      END
    END
  ELSE
    r := rand()
    target := peers[r]
  END
END
         ]]></artwork>
     </figure>
       <t>
         The procedure to determine if we are the closest know peer for a given
         message key and bloomfilter is defined as follows:
       </t>
       <figure anchor="find-bucket">
         <artwork name="" type="" align="left" alt=""><![CDATA[
FIND-BUCKET(peerID, key, kbuckets)
  N := MATCHING-BITS (peerID, key)
  return Nth bucket FROM kbuckets
END
          ]]></artwork>
      </figure>
      <t>The FIND-BUCKET Procedure.</t>
      <figure>
         <artwork name="" type="" align="left" alt=""><![CDATA[
AM-CLOSEST-PEER(key, peerID, bloomfilter, buckets)
  closestPeersBucket := FIND-BUCKET (myPeerID, key, buckets)
  IF key == myPeerID
    return TRUE
  END
  myDistance := XOR(peerID, key)
  FOR EACH p IN closestPeersBucket
    IF XOR(p, key) < myDistance
      return FALSE
    END
    XOR(p, key) == myDistance
      return TRUE
    END
  END
  return TRUE
END
         ]]></artwork>
     </figure>
     <t>The AM-CLOSEST-PEER Procedure.</t>


     </section>
   </section>
   <section anchor="p2p_messages" numbered="true" toc="default">
     <name>Message Processing</name>
      <section anchor="p2p_bf" numbered="true" toc="default">
        <name>Bloomfilter</name>
        <t>
          In order to prevent circular routes, GET and PUT messages contain
          a 128-bit Bloom filter (m=128). The Bloom filter is used to detect duplicate
          peer IDs along the route.
          A Bloom filter "bf" is initially empty, consisting only of zeroes.
          There are two functions which can be invoked on the Bloom filter:
          BF-SET(bf, e) and BF-TEST(bf, e) where "e" is an element which is to added
          to the Bloom filter or queried against the set.
          Any bloom filter uses k=16 different hash functions each of which is
          defined as follows:
        </t>
       <figure>
         <artwork name="" type="" align="left" alt=""><![CDATA[
BF-TEST(key, bloomfilter)
  H_key := SHA512 (key) as UINT32[]
  FOR i IN 0..15
    bit := H_key[i] % 1024
    IF bloomfilter[bit] IS SET
      RETURN TRUE
    END
  END
  RETURN FALSE
END

BF-SET(key, bloomfilter)
  H_key := SHA512 (key) as UINT32[]
  FOR i IN 0..15
    bit := H_key[i] % 1024
    bloomfilter[bit] := 1
  END
END
         ]]></artwork>
     </figure>


      </section>
      <section anchor="p2p_opts" numbered="true" toc="default">
        <name>Processing options</name>
        <t>
          In order to indicate certain processing requirements for messages
          a number of processing options may be specificied in the respective
          field of the signalling messages.
          The options field is 8 octets in length and each options is encoded
          in a single bit.
        </t>
       <dl>
         <dt>Demultiplex everywhere (0)</dt>
         <dd>
           Each peer along the way should process the request. Otherwise
           only peers that are locally closest to the key and no longer in the
           random path mode should process it.
         </dd>
         <dt>Record route (1)</dt>
         <dd>
           Indicates to keep track of the route that the message
           took in the P2P network.
         </dd>
         <dt>Find peer (2)</dt>
         <dd>
           Indicates a 'FIND-PEER' request. Implies that approximate results are
           acceptable.
         </dd>
       </dl>
      </section>
      <section anchor="p2p_xq" numbered="true" toc="default">
        <name>Extended query</name>
        <t>TODO: What is this for? Not documented anywhere</t>
      </section>
     <section anchor="p2p_put" numbered="true" toc="default">
       <name>PUT message</name>
       <section anchor="p2p_put_wire">
         <name>Wire Format</name>
     <figure anchor="figure_putmsg">
       <artwork name="" type="" align="left" alt=""><![CDATA[
0     8     16    24    32    40    48    56
+-----+-----+-----+-----+-----+-----+-----+-----+
|  MSIZE    |   MTYPE   |         BTYPE         |
+-----+-----+-----+-----+-----+-----+-----+-----+
|  OPTIONS  | HOPCOUNT  | REPL_LVL  | PATH_LEN  |
+-----+-----+-----+-----+-----+-----+-----+-----+
|                    EXPIRATION                 |
+-----+-----+-----+-----+-----+-----+-----+-----+
|                   BLOOMFILTER                 /
/                 (128 byte)                    |
+-----+-----+-----+-----+-----+-----+-----+-----+
|                      KEY                      /
/                 (64 byte)                     |
+-----+-----+-----+-----+-----+-----+-----+-----+
/              PUTPATH (variable length)        /
+-----+-----+-----+-----+-----+-----+-----+-----+
/              BLOCK (variable length)        /
+-----+-----+-----+-----+-----+-----+-----+-----+
         ]]></artwork>
     </figure>
     <t>where:</t>
     <dl>
       <dt>MSIZE</dt>
       <dd>
         denotes the size of this message in network byte order.
       </dd>
       <dt>MTYPE</dt>
       <dd>
         is the 16-bit message type. This type can be one of the DHT message
         types but for put messages it must be set to
         the value 146 in network byte order.
       </dd>
       <dt>BTYPE</dt>
       <dd>
         is a 32-bit block type field. The block type indicates the content
         type of the payload. In network byte order.
       </dd>
       <dt>OPTIONS</dt>
       <dd>
         is a 16-bit options field (see below).
       </dd>
       <dt>HOPCOUNT</dt>
       <dd>
         is a 16-bit number indicating how many hops this message has
         traversed to far. In network byte order.
       </dd>
       <dt>REPL_LVL</dt>
       <dd>
         is a 16-bit number indicating the desired replication level of
         the data. In network byte order.
       </dd>
       <dt>PATH_LEN</dt>
       <dd>
         is a 16-bit number indicating the length of the PUT path recorded
         in PUTPATH. As PUTPATH is optiona, this value may be zero.
         In network byte order.
       </dd>
       <dt>EXPIRATION</dt>
       <dd>
         denotes the absolute 64-bit expiration date of the content.
         In microseconds since midnight (0 hour), January 1, 1970 in network
         byte order.
       </dd>
       <dt>BLOOMFILTER</dt>
       <dd>
         A bloomfilter (for peer identities) to stop circular routes.
       </dd>
       <dt>KEY</dt>
       <dd>
         The key under which the PUT request wants to store content
         under.
       </dd>
       <dt>PUTPATH</dt>
       <dd>
         the variable-length PUT path.
         The path consists of a list of PATH_LEN peer IDs.
       </dd>
       <dt>BLOCK</dt>
       <dd>
         the variable-length block payload. The contents are determined
         by the BTYPE field.
       </dd>
     </dl>
   </section>
     <section anchor="p2p_put_processing">
       <name>Processing</name>
       <t>
         Upon receiving a PUT message from a connected peer. An implementation
         MUST process it step by step as follows:
       </t>
       <ol>
         <li>
           The EXPIRATION field is evaluated. If the message is expired,
           it MUST be discarded.
         </li>
         <li>
           If the BTYPE is not supported by the implementation, no validation
           of the block payload is performed and processing continues at (4).
           Else, the block MUST be validated as defined in (3).
         </li>
         <li>
           The block key is extracted from BLOCK. If the block key
           does not match KEY or cannot be extracted because the BLOCK
           is malformed, the message MUST be discarded.
           The block is evaluated. TODO FIXME: In the code, we do not really
           do this. We should review.
         </li>
         <li>
           The sender peer ID SHOULD be in the BLOOMFILTER. If not, the
           implementation MAY log an error, but MUST continue.
         </li>
         <li>
           If the "Record Route" flag is set in OPTIONS, add the local peer ID
           to PUTPATH. FIXME: Should should come way later (?)
         </li>
         <li>
           If the local peer is the closest peer (AM-CLOSEST-PEER) or the
           "Demultiplex Everywhere" options flag ist set, the message MUST
           be stored locally in the block storage.
         </li>
         <li>
           Given the value in REPL_LVL, the number of peers to forward to
           MUST be calculated (NUM-FORWARD-PEERS). If there is at least one
           peer to forward to, the implementation SHOULD select up to this
           number of peers to forward the message to. The implementation MAY
           forward to fewer or no peers in order to handle resource constraints
           such as bandwidth.
           The message BLOOMFILTER MUST be updated with the local peer ID.
         </li>
       </ol>
     </section>
     </section>
     <section anchor="p2p_get" numbered="true" toc="default">
       <name>GET Message</name>
       <section anchor="p2p_get_wire">
         <name>Wire Format</name>
         <figure anchor="figure_getmsg">
         <artwork name="" type="" align="left" alt=""><![CDATA[
0     8     16    24    32    40    48    56
+-----+-----+-----+-----+-----+-----+-----+-----+
|  MSIZE    |   MTYPE   |         BTYPE         |
+-----+-----+-----+-----+-----+-----+-----+-----+
|  OPTIONS  |  HOPCOUNT | REPL_LVL  |  XQ_SIZE  |
+-----+-----+-----+-----+-----+-----+-----+-----+
|                 BLOOMFILTER                   /
/                 (128 byte)                    |
+-----+-----+-----+-----+-----+-----+-----+-----+
|                       KEY                     /
/                 (64 byte)                     |
+-----+-----+-----+-----+-----+-----+-----+-----+
/     BF_MUTATOR        |   XQUERY              /
+-----+-----+-----+-----+                       /
/                 (variable length)             /
+-----+-----+-----+-----+-----+-----+-----+-----+
/              BF_RESULT (variable length)      /
+-----+-----+-----+-----+-----+-----+-----+-----+
         ]]></artwork>
         </figure>
         <t>where:</t>
         <dl>
           <dt>MSIZE</dt>
           <dd>
             denotes the size of this message in network byte order.
           </dd>
           <dt>MTYPE</dt>
           <dd>
             is the 16-bit message type. This type can be one of the DHT message
             types but for put messages it must be set to
             the value 147 in network byte order.
           </dd>
           <dt>BTYPE</dt>
           <dd>
             is a 32-bit block type field. The block type indicates the content
             type of the payload. In network byte order.
           </dd>
           <dt>OPTIONS</dt>
           <dd>
             is a 16-bit options field (see below).
           </dd>
           <dt>HOPCOUNT</dt>
           <dd>
             is a 16-bit number indicating how many hops this message has
             traversed to far. In network byte order.
           </dd>
           <dt>REPL_LVL</dt>
           <dd>
             is a 16-bit number indicating the desired replication level of
             the data. In network byte order.
           </dd>
           <dt>XQ_SIZE</dt>
           <dd>
             is a 32-bit number indicating the length of the optional
             extended query XQUERY. In network byte order.
           </dd>
           <dt>BLOOMFILTER</dt>
           <dd>
             A bloomfilter (for peer identities) to stop circular routes.
           </dd>
           <dt>KEY</dt>
           <dd>
             The key under which the PUT request wants to store content
             under.
           </dd>
           <dt>XQUERY</dt>
           <dd>
             the variable-length extended query. Optional.
           </dd>
           <dt>BF_MUTATOR</dt>
           <dd>
             The 32-bit bloomfilter mutator for the result bloomfilter.
           </dd>
           <dt>RESULT_BF</dt>
           <dd>
             the variable-length result bloomfilter.
           </dd>
         </dl>
       </section>
       <section anchor="p2p_get_processing">
         <name>Processing</name>
         <t>
           Upon receiving a GET message from a connected peer. An implementation
           MUST process it step by step as follows:
         </t>
         <ol>
           <li>
             The KEY and XQUERY is validated against the requested BTYPE.
             If the BTYPE is not supported, or if the block key
             does not match the BTYPE or if the XQUERY is malformed,
             the message MUST be discarded.
           </li>
           <li>
             The sender peer ID SHOULD be in the BLOOMFILTER. If not, the
             implementation MAY log an error, but MUST continue.
           </li>
           <li>
             <t>
               If the local peer is the closest peer (AM-CLOSEST-PEER) or the
               "Demultiplex Everywhere" options flag is set, a reply MUST be
               produced:
             </t>
             <ol>
               <li>
                 If OPTIONS indicate a "Find Peer" request, FIXME the peer selection
                 foo from buckets that probably needs fixing. Take into account
                 REPLY_BF
               </li>
               <li>
                 Else, if there is a BLOCK in the local Block Storage which is
                 not already in the RESULT_BF, a RESULT message MUST be sent.
                 FIXME link to how the result is sent?
               </li>
             </ol>
           </li>
           <li>
             FIXME: We only handle if not GNUNET_BLOCK_EVALUATION_OK_LAST??
           </li>
           <li>
             Given the value in REPL_LVL, the number of peers to forward to
             MUST be calculated (NUM-FORWARD-PEERS). If there is at least one
             peer to forward to, the implementation SHOULD select up to this
             number of peers to forward the message to. The implementation MAY
             forward to fewer or no peers in order to handle resource constraints
             such as bandwidth.
             The message BLOOMFILTER MUST be updated with the local peer ID.
           </li>
         </ol>
       </section>
     </section>
     <section anchor="p2p_result" numbered="true" toc="default">
       <name>RESULT message</name>
       <section anchor="p2p_result_wire">
         <name>Wire Format</name>
         <figure anchor="figure_resmsg">
           <artwork name="" type="" align="left" alt=""><![CDATA[
0     8     16    24    32    40    48    56
+-----+-----+-----+-----+-----+-----+-----+-----+
|  MSIZE    |   MTYPE   |        BTYPE          |
+-----+-----+-----+-----+-----+-----+-----+-----+
|   //      | OPTIONS   | PUTPATH_L | GETPATH_L |
+-----+-----+-----+-----+-----+-----+-----+-----+
|                   EXPIRATION                  |
+-----+-----+-----+-----+-----+-----+-----+-----+
|                      KEY                      /
/                 (64 byte)                     |
+-----+-----+-----+-----+-----+-----+-----+-----+
/                    PUTPATH                    /
/                 (variable length)             /
+-----+-----+-----+-----+-----+-----+-----+-----+
/                    GETPATH                    /
/                 (variable length)             /
+-----+-----+-----+-----+-----+-----+-----+-----+
/                   BLOCK                       /
/              (variable length)                /
+-----+-----+-----+-----+-----+-----+-----+-----+
         ]]></artwork>
         </figure>
         <t>where:</t>
         <dl>
           <dt>MSIZE</dt>
           <dd>
             denotes the size of this message in network byte order.
           </dd>
           <dt>MTYPE</dt>
           <dd>
             is the 16-bit message type. This type can be one of the DHT message
             types but for put messages it must be set to
             the value 148 in network byte order.
           </dd>
           <dt>OPTIONS</dt>
           <dd>
             is a 16-bit options field (see below).
           </dd>
           <dt>BTYPE</dt>
           <dd>
             is a 32-bit block type field. The block type indicates the content
             type of the payload. In network byte order.
           </dd>
           <dt>PUTPATH_L</dt>
           <dd>
             is a 16-bit number indicating the length of the PUT path recorded
             in PUTPATH. As PUTPATH is optiona, this value may be zero.
             In network byte order.
           </dd>
           <dt>GET_PATH_LEN</dt>
           <dd>
             is a 16-bit number indicating the length of the GET path recorded
             in GETPATH. As PUTPATH is optiona, this value may be zero.
             In network byte order.
           </dd>
           <dt>EXPIRATION</dt>
           <dd>
             denotes the absolute 64-bit expiration date of the content.
             In microseconds since midnight (0 hour), January 1, 1970 in network
             byte order.
           </dd>
           <dt>KEY</dt>
           <dd>
             The key under which the PUT request wants to store content
             under.
           </dd>
           <dt>PUTPATH</dt>
           <dd>
             the variable-length PUT path.
             The path consists of a list of PATH_LEN peer IDs.
           </dd>
           <dt>GETPATH</dt>
           <dd>
             the variable-length PUT path.
             The path consists of a list of PATH_LEN peer IDs.
           </dd>
           <dt>BLOCK</dt>
           <dd>
             the variable-length resource record data payload.
             The contents are defined by the respective type of the resource record.
           </dd>
         </dl>
       </section>
       <section anchor="p2p_result_processing">
         <name>Processing</name>
         <t>
           Upon receiving a RESULT message from a connected peer. An implementation
           MUST process it step by step as follows:
         </t>
         <ol>
           <li>
             The EXPIRATION field is evaluated. If the message is expired,
             it MUST be discarded.
           </li>
           <li>
             If the MTYPE of the message indicates a HELLO block, the
             payload MUST be considered for the local routing table.
             FIXME: Considered how?
           </li>
           <li>
             If the sender peer (FIXME which peer?) is already found in the
             GETPATH, the path MUST be truncated.
           </li>
           <li>
             If the KEY of this PUT message is found in the list of pending
             queries, the the KEY and XQUERY is validated against the requested
             BTYPE.
             If the BTYPE is not supported, or if the block key
             does not match the BTYPE or if the XQUERY is malformed,
             the message MUST be discarded. (FIXME: It is not clear the key
             validation is happening. However, block validation is.)
           </li>
           <li>
             The implementation MAY cache RESULT messages.
           </li>
           <li>
             If no requests for this KEY or BTYPE are known, result processing
             is completed.
           </li>
           <li>
             If the request is of type "Find Peer" and the message BTYPE is
             of type HELLO the block key is extracted from BLOCK, and if the
             block key does not match KEY or cannot be extracted because
             the BLOCK is malformed, the message MUST be discarded.
             Otherwise, the block is evaluated against the message KEY.
             FIXME: If OK_MORE or OK_LAST the RESULT is routed. One (!) peer is
             selected from the connected peers (!). If none is found the message
             is discarded.
           </li>
         </ol>
       </section>
     </section>
   </section>
   <section anchor="blockstorage" numbered="true" toc="default">
     <name>Block Storage</name>
     <section>
       <name>Block Processing</name>
       <t>RequestEvaluationResult</t>
       <dl>
         <dt>REQUEST_VALID</dt>
         <dd>Query is valid, no reply given.</dd>
         <dt>REQUEST_INVALID</dt>
         <dd>
           Query format does not match block type. For example, XQuery not
           given or of size of XQuery is not appropriate for type.
         </dd>
       </dl>
       <t>ReplyEvaluationResult</t>
       <dl>
         <dt>OK_MORE</dt>
         <dd>Valid result, and there may be more.</dd>
         <dt>OK_LAST</dt>
         <dd>Last possible valid result.</dd>
         <dt>OK_DUPLICATE</dt>
         <dd>Valid result, but duplicate.</dd>
         <dt>RESULT_INVALID</dt>
         <dd>Invalid result. Block does not match query. Value = 4.</dd>
         <dt>RESULT_IRRELEVANT</dt>
         <dd>Block does not match xquery. Valid result, but not relevant for the request.</dd>
       </dl>
     </section>
     <section anchor="block_functions">
       <name>Block Functions</name>
       <t>
         Any block type implementation MUST implement the following functions.
       </t>
       <dl>
         <dt>ValidateBlockQuery(Key, XQuery) -> RequestEvaluationResult</dt>
         <dd>
           is used to evaluate the request for a block. It is used as part of
           <tt>GetMessage</tt> processing, where the block payload is still unkown,
           but the block <tt>XQuery</tt> (FIXME: Undefined here) and <tt>Key</tt> can and
           MUST be verified, if possible.
         </dd>
         <dt>ValidateBlockStoreRequest(Block, Key) -> RequestEvaluationResult</dt>
         <dd>
           is used to evaluate a block including its key and payload.
           It is used as part of <tt>PutMessage</tt> processing.
           The validation MUST include a check of the block payload against the
           <tt>Key</tt> under which it is requested to be stored.
         </dd>
         <dt>ValidateBlockReply(Block, XQuery, Key) -> ReplyEvaluationResult</dt>
         <dd>
           is used to evaluate a block including its <tt>Key</tt> and payload.
           It is used as part <tt>ResultMessage</tt> processing.
           The validation of the respective <tt>Block</tt> requires a pending local query
           or a previously routed request of another peer and its associated
           XQuery data and Key.
           The validation MUST include a check of the block payload against the
           key under which it is requested to be stored.
         </dd>
         <dt>DeriveBlockKey(Block) -> Key</dt>
         <dd>
           is used to synthesize the block key from the block payload
           and metadata. It is used as part of FIND-PEER message processing.
         </dd>
         <dt>FilterResult(Block, XQuery, Key) -> ReplyEvaluationResult</dt>
         <dd>
           is used to filter results stored in the local block storage for local
           queries. Locally stored blocks from previously observed
           <tt>ResultMessages</tt> and <tt>PutMessages</tt> MAY use this
           function instead of <tt>ValidateBlockReply</tt> in order to
           avoid revalidation of the block and only perform filtering based on
           request parameters.
         </dd>
       </dl>
     </section>
     <section>
       <name>Block Types</name>
       <t>
         Applications can and should define their own block types.
         The block type determines the format and handling of the block
         payload by peers in PUT and RESULT messages.
         Block types MUST be registered with GANA <xref target="gana"/>.
       </t>
       <t>
         For bootstrapping and peer discovery, the DHT implementation uses
         its own block type called "HELLO". A block with this block type
         contains the peer ID of the peer initiating the GET request.
       </t>
       <section>
         <name>HELLO</name>
         <t>
           The HELLO block type wire format is illustrated in
           <xref target="figure_hellobt"/>. A block of type HELLO MUST NOT
           include extended query data (xquery). Any implementation
           encountering a HELLO block with xquery data MUST consider the
           block invalid and ignore it.
         </t>
         <t>
           A HELLO reply block MAY be empty. Otherwise, it contains the
           HELLO URI of a peer.
         </t>
         <figure anchor="figure_hellobt">
           <artwork name="" type="" align="left" alt=""><![CDATA[
         FIXME: Wire format
         ]]></artwork>
         </figure>
       </section>
     </section>
   </section>


   <section>
     <name>Bootstrapping</name>
     <t>
       It is assumed that the peer is already connected to at least
       one other peer.
       First, those initial peers are sorted into their respective buckets.
     </t>
     <t>
       In order to find the closest peers in the network to itself, an
       implementation MUST now periodically send HELLO GET queries for its own
       peer ID.
       Both the "record route" and "find peer" message options are set in the
       GET queries in order to learn peers and network topology from the
       message route and in order to receive approximate replies to the
       query key (the peer ID).
     </t>
     <t>FIXME: Periodically -> more specific? No. Frequency may be adapted depending on network conditions, known peers, busy/idle etc.</t>
     <t>
       Any implementation encountering a HELLO GET request initially
       sends its own peer ID if it.
     </t>
   </section>
   <section anchor="security" numbered="true" toc="default">
     <name>Security Considerations</name>
     <!-- FIXME: Here we should (again) discuss how the system is open and
     does not have/require a trust anchor a priori. This is (again) in contrast
     to RELOAD -->
   </section>
   <section anchor="gana" numbered="true" toc="default">
       <name>GANA Considerations</name>
       <t>
         GANA <xref target="GANA" />
         is requested to create a "DHT Block Types" registry.
         The registry shall record for each entry:
       </t>
       <ul>
         <li>Name: The name of the block type (case-insensitive ASCII
           string, restricted to alphanumeric characters</li>
         <li>Number: 32-bit</li>
         <li>Comment: Optionally, a brief English text describing the purpose of
           the block type (in UTF-8)</li>
         <li>Contact: Optionally, the contact information of a person to contact for
           further information</li>
         <li>References: Optionally, references describing the record type
           (such as an RFC)</li>
       </ul>
       <t>
         The registration policy for this sub-registry is "First Come First
         Served", as described in <xref target="RFC8126"/>.
         GANA is requested to populate this registry as follows:
       </t>
       <figure anchor="figure_btypenums">
         <artwork name="" type="" align="left" alt=""><![CDATA[
Number | Name   | Contact | References | Description
-------+--------+---------+------------+-------------------------
0       ANY      N/A       [This.I-D]   Reserved
7       HELLO    N/A       [This.I-D]   Type of a block that contains
                                        a HELLO for a peer
11      GNS      N/A       GNS          Block for storing record data
           ]]></artwork>
       </figure>
       <t>
         GANA is requested to amend the "GNUnet Signature Purpose" registry
         as follows:
       </t>
       <figure anchor="figure_purposenums">
         <artwork name="" type="" align="left" alt=""><![CDATA[
Purpose | Name            | References | Description
--------+-----------------+------------+--------------------------
           ]]></artwork>
       </figure>
     </section>
     <!-- gana -->
     <section>
       <name>Test Vectors</name>
   </section>
   </middle>
   <back>
     <references>
       <name>Normative References</name>

         &RFC2119;
         &RFC3629;
         &RFC3986;
         &RFC4648;
         &RFC6940;
         &RFC8126;
         &RFC8174;
 
      <reference anchor="ed25519" target="http://link.springer.com/chapter/10.1007/978-3-642-23951-9_9">
         <front>
           <title>High-Speed High-Security Signatures</title>
          <author initials="D." surname="Bernstein" fullname="Daniel Bernstein">
            <organization>University of Illinois at Chicago</organization>
          </author>

          <author initials="N." surname="Duif"
            fullname="Niels Duif">
          <organization>Technische Universiteit Eindhoven</organization>

        </author>
          <author initials="T." surname="Lange"
            fullname="Tanja Lange">
          <organization>Technische Universiteit Eindhoven</organization>

          </author>
          <author initials="P." surname="Schwabe"
            fullname="Peter Schwabe">
          <organization>National Taiwan University</organization>

          </author>
          <author initials="B." surname="Yang"
            fullname="Bo-Yin Yang">
          <organization>Academia Sinica</organization>

          </author>
           <date year="2011"/>
         </front>
       </reference>

       <reference anchor="CrockfordB32" target="https://www.crockford.com/base32.html">
         <front>
           <title>Base32</title>
          <author initials="D." surname="Douglas" fullname="Crockford">
          </author>

           <date year="2019" month="March"/>
         </front>
       </reference>

       <reference anchor="GANA" target="https://gana.gnunet.org/">
         <front>
           <title>GNUnet Assigned Numbers Authority (GANA)</title>
           <author><organization>GNUnet e.V.</organization>
           </author>
           <date month="April" year="2020" />
         </front>
       </reference>



     </references>
     <references>
       <name>Informative References</name>
      <reference anchor="R5N" target="https://doi.org/10.1109/ICNSS.2011.6060022">
         <front>
           <title>R5N: Randomized recursive routing for restricted-route networks</title>
          <author initials="N. S." surname="Evans" fullname="Nathan S. Evans">
            <organization>Technische Universität München</organization>
          </author>

          <author initials="C." surname="Grothoff"
            fullname="Christian Grothoff">
          <organization>Technische Universität München</organization>
          </author>
           <date year="2011"/>
         </front>
       </reference>
       <reference anchor="Kademlia" target="http://css.csail.mit.edu/6.824/2014/papers/kademlia.pdf">
         <front>
           <title>Kademlia: A peer-to-peer information system based on the xor metric.</title>
          <author initials="P." surname="Maymounkov" fullname="Petar Maymounkov">
          </author>

          <author initials="D." surname="Mazieres"
            fullname="David Mazieres">
        </author>
           <date year="2002"/>
         </front>
       </reference>

       <reference anchor="cadet" target="https://doi.org/10.1109/MedHocNet.2014.6849107">
         <front>
           <title>CADET: Confidential ad-hoc decentralized end-to-end transport</title>
          <author initials="B." surname="Polot" fullname="Bartlomiej Polot">
            <organization>Technische Universität München</organization>
          </author>

          <author initials="C." surname="Grothoff"
            fullname="Christian Grothoff">
          <organization>Technische Universität München</organization>
          </author>
           <date year="2014"/>
         </front>
       </reference>
       <reference anchor="I-D.draft-schanzen-gns" target="https://datatracker.ietf.org/doc/draft-schanzen-gns/">
         <front>
           <title>The GNU Name System</title>
          <author initials="M." surname="Schanzenbach" fullname="Martin Schanzenbach">
            <organization>GNUnet e.V.</organization>
          </author>

          <author initials="C." surname="Grothoff"
            fullname="Christian Grothoff">
          <organization>GNUnet e.V.</organization>
        </author>
          <author initials="B." surname="Fix"
            fullname="Bernd Fix">
          <organization>GNUnet e.V.</organization>
          </author>
           <date year="2021"/>
         </front>
       </reference>



     </references>
     <!-- Change Log
       v00 2017-07-23  MS   Initial version
     -->
   </back>
 </rfc>