One document matched: draft-bryan-p2psip-reload-03.xml
<?xml version="1.0" encoding="US-ASCII"?>
<!DOCTYPE rfc SYSTEM "rfc2629.dtd">
<?xml-stylesheet type="text/xsl" href="rfc2629.xslt" ?>
<?rfc toc="yes" ?>
<?rfc symrefs="yes" ?>
<?rfc iprnotified="no" ?>
<?rfc strict="no" ?>
<?rfc compact="yes" ?>
<?rfc sortrefs="no" ?>
<?rfc colonspace="yes" ?>
<?rfc rfcedstyle="no" ?>
<rfc category="std" docName="draft-bryan-p2psip-reload-03" ipr="full3978">
<front>
<title abbrev="RELOAD">REsource LOcation And Discovery (RELOAD)</title>
<author fullname="Cullen Jennings" initials="C." surname="Jennings">
<organization>Cisco</organization>
<address>
<postal>
<street>170 West Tasman Drive</street>
<street>MS: SJC-21/2</street>
<city>San Jose</city>
<region>CA</region>
<code>95134</code>
<country>USA</country>
</postal>
<phone>+1 408 421-9990</phone>
<email>fluffy@cisco.com</email>
</address>
</author>
<author fullname="Bruce B. Lowekamp" initials="B. B." surname="Lowekamp">
<organization>SIPeerior; William & Mary</organization>
<address>
<postal>
<street>3000 Easter Circle</street>
<city>Williamsburg</city>
<region>VA</region>
<code>23188</code>
<country>USA</country>
</postal>
<phone>+1 757 565 0101</phone>
<email>lowekamp@sipeerior.com</email>
</address>
</author>
<author fullname="Eric Rescorla" initials="E.K." surname="Rescorla">
<organization>Network Resonance</organization>
<address>
<postal>
<street>2064 Edgewood Drive</street>
<city>Palo Alto</city>
<region>CA</region>
<code>94303</code>
<country>USA</country>
</postal>
<phone>+1 650 320-8549</phone>
<email>ekr@networkresonance.com</email>
</address>
</author>
<author fullname="Jonathan Rosenberg" initials="J.D." surname="Rosenberg">
<organization>Cisco</organization>
<address>
<postal>
<street></street>
<city>Edison</city>
<region>NJ</region>
<country>USA</country>
</postal>
<email>jdrosen@cisco.com</email>
</address>
</author>
<author fullname="Salman A. Baset" initials="S.A." surname="Baset">
<organization>Columbia University</organization>
<address>
<postal>
<street>1214 Amsterdam Avenue</street>
<city>New York</city>
<region>NY</region>
<country>USA</country>
</postal>
<email>salman@cs.columbia.edu</email>
</address>
</author>
<author fullname="Henning Schulzrinne" initials="H.G."
surname="Schulzrinne">
<organization>Columbia University</organization>
<address>
<postal>
<street>1214 Amsterdam Avenue</street>
<city>New York</city>
<region>NY</region>
<country>USA</country>
</postal>
<email>hgs@cs.columbia.edu</email>
</address>
</author>
<date day="24" month="February" year="2008" />
<area>RAI</area>
<workgroup>P2PSIP</workgroup>
<abstract>
<t>This document defines REsource LOcation And Discovery (RELOAD), a
peer-to-peer (P2P) binary signaling protocol for use on the Internet. A
P2P signaling protocol provides its clients with an abstract hash table
service between a set of cooperating peers that form the overlay
network. RELOAD is designed to support a P2P Session Initiation Protocol
(P2PSIP) network, but can be utilized by other applications with similar
requirements by defining new usages that specify the data kinds that
must be stored for a particular application. RELOAD defines a security
model based on a certificate enrollment service that provides unique
identities. NAT traversal is a fundamental service of the protocol.
RELOAD also allows access from "client" nodes which do not need to route
traffic or store data for others.</t>
</abstract>
</front>
<middle>
<section title="Introduction">
<t>This document defines REsource LOcation And Discovery (RELOAD), a
peer-to-peer (P2P) signaling protocol for use on the Internet. It
provides a Distributed Hash Table (DHT) service, which allows
participating nodes to read and write entries into a hash table that is
stored collectively among the participants. RELOAD is a lightweight,
binary protocol. It provides several functions that are critical for a
successful P2P protocol for the Internet. These are:</t>
<t><list style="hanging">
<t hangText="Security Framework:">Security is one of the most
challenging problems in a P2P protocol. A P2P network will often be
established among a set of peers that do not trust each other. Yet,
despite this lack of trust, the network must operate reliably to
allow storage and retrieval of data. RELOAD defines an abstract
enrollment server, which all entities trust to generate unique
identifiers for each user. Using that small amount of trust as an
anchor, RELOAD defines a security framework that allows for
authorization of P2P protocol functions and authentication of data
stored in the overlay. This does not remove all attacks but greatly
reduces the possible attack space.</t>
<t hangText="Usage Model:">RELOAD is designed to support a variety
of applications, including P2P multimedia communications with the
Session Initiation Protocol <xref
target="I-D.ietf-p2psip-concepts"></xref>. Consequently, RELOAD has
the notion of a usage, one of which is defined to support each
application (this document also defines the SIP usage for multimedia
communications). Each usage identifies a set of data kind that need
to be stored and retrieved from the DHT. Each kind defines a data
structure, authorization policies, size quota, and information
required for storage and retrieval in the DHT. The usage concept
allows RELOAD to be used with new applications through a simple
documentation process that supplies the details for each
application.</t>
<t hangText="NAT Traversal:">Operations for NAT traversal are part
of the base design, including establishing new RELOAD connections
and tunneling SIP or other application protocols required by P2PSIP.
RELOAD makes use of Interactive Connectivity Establishment (ICE)
<xref target="I-D.ietf-mmusic-ice"></xref> to facilitate the
creation of the P2P network and the establishment of channels for
use by the application protocol (SIP and RTP, for example). RELOAD
also defines how peers in the P2P network act as STUN and TURN
servers and how those resources can be discovered through the DHT.
With these features, RELOAD can run in modes in which nearly all the
peers are behind NATs, yet are able to fully participate without
imposing any constraints on the actual DHT algorithm.</t>
<t hangText="High Performance Routing:">The very nature of DHT
algorithms introduces a requirement that peers participating in the
P2P network route requests on behalf of other peers in the network.
This introduces a load on those other peers, in the form of
bandwidth and processing power. RELOAD has been defined to reduce
the amount of bandwidth and processing required of peers. It does so
by using a very lightweight binary protocol, and furthermore, by
defining a packet structure that facilitates low-complexity
forwarding, including hardware-based forwarding. In particular, a
generic transport header is used for routing the message through the
overlay without the contents needing to be parsed by (or even
visible to) intermediate peers. The header includes no information
about specific IP addresses because none are needed to route along
an overlay. The header only includes lists of peers which the
message should be routed through/too, as well as some minor options
and version flags. Clearly separating the header components
necessary for routing from the message contents simplifies
processing and increases security.</t>
<t hangText="Transport Flexibility:">RELOAD has native support for
both DTLS and TLS for the underlying transport protocol, with
support for DTLS over UDP as mandatory to implement. TLS over TCP is
preferred because it has better bulk data performance and connection
stability, but UDP is more likely to provide direct connections
between peers in the presence of NATs. Explicit support for
fragmentation is provided and required when using UDP. Because there
is no single universally available and suitable transport protocol,
the peer protocol must be flexible in this regard. New transports
can be supported trivially.</t>
<t hangText="Pluggable DHT Algorithms:">RELOAD has been designed
with an abstract interface to the overlay layer to simplify
implementing a variety of structured (DHT) and unstructured overlay
algorithms. This specification also defines how RELOAD is used with
Chord, which is mandatory to implement. Specifying a default "must
implement" DHT will allow interoperability, while the extensibility
allows selection of DHTs optimized for a particular application.</t>
</list></t>
<t>These properties were designed specifically to meet the requirements
for a P2P protocol to support SIP. However, RELOAD is not limited to
usage by SIP and could serve as a tool for supporting other P2P
applications with similar needs. RELOAD is also based on the concepts
introduced in <xref target="I-D.ietf-p2psip-concepts"></xref>.</t>
<section title="Architecture">
<t>Architecturally this specification is divided into several layers,
as shown in the following figure.</t>
<figure>
<artwork><![CDATA[
Application
-------------------------------------- Usage-defined API
+-------+ +-------+
Usage | SIP | | XMPP | ...
Layer | Usage | | Usage |
+-------+ +-------+
-------------------------------------- Distributed Storage API
Overlay Overlay +-------------+
Routing & Routing & +----+ | +-----+ |
Storage Replication | DB | | |Chord| ... | Topology
Layer Logic +----+ | | | | Plugins
| +-----+ |
+-------------+
--------------------------------------
+------+ +-----+
Forwarding Forwarding & | STUN | | ICE |
Layer Encoding Logic +------+ +-----+
-------------------------------------- Common Packet Encoding
Transport +-------+ +------+
Layer |TLS | |DTLS |
+-------+ +------+
]]></artwork>
</figure>
<t>The three layers defined by RELOAD include:</t>
<t><list style="hanging">
<t hangText="Usage Layer:">Provides an application-specific
interface that maps an application's requirements onto the generic
services of the DHT.</t>
<t hangText="Overlay Routing & Storage Layer:">Implements the
overlay. Chooses what links to establish to form the DHT's overlay
network, manages the storage and migration of data for this peer
and on behalf of other peers, and performs searches for requested
data across the DHT.</t>
<t hangText="Forwarding Layer:">Provides packet forwarding
services between nodes. Also handles setting up connections across
NATs using ICE.</t>
</list></t>
<section title="Usage Layer">
<t>The top layer, called the Usage Layer, has application usages,
such as the SIP Location Usage, that use the abstract distributed
storage API to store and retrieve data from the DHT. The goal of
this layer is to implement application-specific usages of the
Overlay Routing and Storage Layer below it. The Usage defines how a
specific application maps its data into something that can be stored
in the DHT, where to store the data, how to secure the data, and
finally how applications can retrieve and use the data.</t>
<t>The architecture diagram shows both a SIP usage and an XMPP
usage. A single application may require multiple usages. A usage may
define multiple kinds of data that are stored in the overlay and may
also rely on kinds originally defined by other usages. A usage is
not itself encoded on the wire, only the kind-ids and data models
are, but is rather a specification of the functionality that is
required for a given application. That specification typically
specifies semantics, access control rules, and the format and size
of the data which may be stored.</t>
<t>One usage may depend on another. For example, the SIP usage
depends on a Certificate Store usage (not shown in the diagram) to
obtain the certificates required to authenticate messages. Because
certificates are stored in standard X.509 form, there is no reason
for each usage to specify this service independently.</t>
</section>
<section title="Overlay Routing and Storage Layer">
<t>The Overlay Routing and Storage Layer stores and retrieves
information, performs maintenance of the DHT as peers join and leave
the DHT, and routes messages on the overlay. The DHT implementation
is provided by a pluggable component so that each overlay can select
an appropriate DHT that relies on the common RELOAD core code.</t>
<t>The Overlay Routing and Replication Logic provides a fairly
generic interface that allows the DHT implementation to control the
overlay and resource operations and messages. Since each DHT is
defined and functions differently, we generically refer to the table
of other peers that the DHT maintains and uses to route requests
(neighbors) as a Routing Table. The Logic component makes queries to
the DHT's Routing Table to determine the next hop, then encodes and
sends the message itself. Similarly, the DHT issues periodic update
requests through the logic component to maintain and update its
Routing Table.</t>
<t>The DHT shown in the illustration is Chord, but a variety of DHT
algorithms are possible through a pluggable interface. A single node
could be functioning in multiple overlays simultaneously, each using
its own DHT algorithm. Each peer is identified by and its location
in the overlay determined by its Peer-ID that is assigned by the
enrollment server when the user or peer first enrolls in the
overlay. The Peer-ID also determines the set of resources which it
will be responsible for storing. The exact mapping between these is
determined by the DHT algorithm used by the overlay, therefore the
logic component always queries the DHT to determine where a
particular resource should be stored.</t>
<t>As peers enter and leave, resources may be stored on different
peers, so the information related to them is exchanged as peers
enter and leave. Redundancy is used to protect against loss of
information in the event of a peer failure and to protect against
compromised or subversive peers. The Logic component notifies the
DHT as neighbors join and leave, and the DHT updates its Routing
Table and issues resource migration requests as appropriate.</t>
</section>
<section title="Forwarding Layer">
<t>This layer is responsible for getting a packet to the next peer,
as determined by the Routing and Storage Layer. The Forwarding Layer
establishes and maintains the network connections required by the
DHT's Routing Table. This layer is also responsible for setting up
connections to other peers through NATs and firewalls using ICE, and
it can elect to forward traffic using relays for NAT and firewall
traversal.</t>
</section>
</section>
<section title="Security">
<t>RELOAD provides two security mechanisms, one based on public key
certificates and one based on a globally shared key.</t>
<t>RELOAD's preferred security framework is built upon an enrollment
server. The enrollment server issues each new peer a certificate that
assigns it a Peer-ID. By generating the Peer-IDs randomly and
controlling what peers are issued certificates, the enrollment server
protects against many of the attacks on the overlay network.
Similarly, all users are issued certificates for their identities by
the enrollment server. All resources stored on the overlay must be
signed by their creator, thus ensuring that an attacker cannot forge
data belonging to another user. The enrollment process is only
required to join the overlay (and perhaps to refresh an expired
certificate). The peers and users do not need to have real-time access
to the enrollment server.</t>
<t>TLS or DTLS are used for communication between peers. In
combination with the certificates, this provides both confidentiality
and authentication for communication across the overlay. Applications
such as P2PSIP can also make use of the users' certificates to achieve
secure end-to-end connections at the application layer.</t>
<t>RELOAD also provides for an authorization framework based on
certificates. Each usage defines the access control rules for which
users/peers are allowed to read/write each Resource-ID. This protects
data belonging to one user from being written by another, as well as
allowing for a distributed quota mechanism based on the maximum size
of each kind of data.</t>
<!-- <t>Further security is provided by replicating the resources stored
on the overlay. Even if an attacker can compromise a single peer
(either directly or through the routing), by locating replicas
randomly around the overlay, a valid copy of the data is likely to
be obtained. </t>-->
<t>In addition to the enrollment server model, RELOAD offers a
security model using a pre-shared-key. Although this provides
significantly less security than is provided through an enrollment
server, it allows ad hoc or ephemeral overlays to be set up with
minimal effort on the part of the users.</t>
</section>
</section>
<section title="Terminology">
<t>The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in <xref
target="RFC2119">RFC 2119</xref>.</t>
<t>We use the terminology and definitions from the <xref
target="I-D.ietf-p2psip-concepts">Concepts and Terminology for Peer to
Peer SIP</xref> draft extensively in this document. Other terms used in
this document are defined inline when used and are also defined below
for reference.</t>
<t>The following important terms from the Concepts document are defined
below for reference.</t>
<t><list style="hanging">
<t hangText="DHT:">A distributed hash table. A DHT is an abstract
hash table service realized by storing the contents of the hash
table across a set of peers.</t>
<t hangText="DHT Algorithm:">A DHT algorithm defines the rules for
determining which peers in a DHT store a particular piece of data
and for determining a topology of interconnections amongst peers in
order to find a piece of data.</t>
<t hangText="DHT Instance:">A specific hash table and the collection
of peers that are collaborating to provide read and write access to
it. There can be any number of DHT instances running in an IP
network at a time, and each operates in isolation of the others.</t>
<t hangText="P2P Network:">Another name for a DHT instance.</t>
<t hangText="P2P Network Name:">A string that identifies a unique
P2P network. P2P network names are DNS names - for example,
"example.org". Lookup of such a name in DNS returns services
associated with the DHT, such as enrollment servers, bootstrap
peers, or gateways (for example, a SIP gateway between a traditional
SIP and a P2P SIP network called "example.com").</t>
<t hangText="Resource-ID:">A value that is not human friendly to
read and identifies some resources and which is used as a key for
storing and retrieving the resource. One way to generate a
Resource-ID is by applying a mapping function to some other unique
name (e.g., user name or service name) for the resource. The
Resource-ID is used by the distributed database algorithm to
determine the peer or peers that are responsible for storing the
data for the overlay. In structured P2P networks, resource-IDs are
generally fixed length and are formed by hashing the resource
identifier. In unstructured networks, resource identifiers may be
used directly as resource-IDs and may have variable length.</t>
<t hangText="Peer:">A host that is participating in the DHT. By
virtue of its participation it can store data and is responsible for
some portion of the overlay.</t>
<t hangText="Peer-ID:">A value that uniquely identifies a peer.
Peer-IDs 0 and 2^N - 1 are reserved and are invalid peer-IDs. A
value of zero is not used in the wire protocol but can be used to
indicate an invalid peer in implementations and APIs. The peer-id of
2^N-1 is used on the wire protocol as a wildcard.</t>
<t hangText="Resource:">An object associated with a string
identifier. In unstructured P2P networks, the identifier is used
directly as a Resource-Id. In structured P2P networks the identifier
can be mapped into a Resource-ID by using the string as the input to
the hash function. A SIP resource, for example, is identified by its
AOR.</t>
<t hangText="User:">A human being.</t>
</list>></t>
<t>We also introduce the following important new terms.</t>
<t><list style="hanging">
<t hangText="Connection Table:">The set of peers to which a peer is
directly connected. This includes peers with which CONNECT
handshakes have been done but which have not sent any UPDATEs.</t>
<t hangText="Routing Table:">The set of peers which a peer can use
to route DHT messages. In general, these peers will all be on the
connection table but not vice versa, because some peers will have
CONNECTed but not sent updates. Peers may send messages directly to
peers which are on the connection table but may only route messages
to other peers through peers which are on the routing table.</t>
<t hangText="Hashed-ID:">The generic term for an identifier in the
hash space of the DHT. Examples of Hashed-IDs include Resource-IDs
and Peer-IDs. This only applies to structured overlays.</t>
<t hangText="Unhashed-ID:">An Unhashed-ID is a string used as an
input to a hash function, the result of which is a Hashed-ID. This
only applies to structured overlays.</t>
<t hangText="Usage:">A usage is an application that wishes to use
the DHT for some purpose. Each application wishing to use the DHT
defines a set of data kinds that it wishes to use. The SIP usage
defines the location, certificate, STUN server and TURN server data
kinds.</t>
<t hangText="Destination List:">A list of IDs through which a
message is to be routed. This allows for request/response source
routing. A single ID is a trivial form of destination list.</t>
</list></t>
</section>
<section title="Overview">
<section title="Distributed Storage Layer">
<t>RELOAD is designed to be extensible to both structured and
unstructured overlays. However, this version is only completely worked
out for structured overlays such as DHTs. The following text assumes
structured overlays; in particular Resource-IDs are assumed to be
fixed length for any given overlay, although the protocol allows them
to be variable length to allow extension to unstructured overlays.</t>
<t>Each logical address in the DHT where data can be stored is
referred to as a Resource-ID. A given peer will be responsible for
storing data from many Resource-ID locations. Typically literature on
DHTs uses the term "key" to refer to a location in the DHT; however,
in this specification the term key is used to refer to public or
private keys used for cryptographic operations and the term
Resource-ID is used to refer to a location for storage in the DHT.</t>
<section title="DHT Concepts">
<t>While very early P2P systems used flood based techniques, some
newer P2P systems locate resources using a Distributed Hash Table,
or DHT to improve efficiency. Peers are organized using a
Distributed Hash Table (DHT) structure. In such a system, every
resource has a Resource-ID, which is obtained by hashing some
keyword or value (an Unhashed-ID) that uniquely identifies the
resource. Resources can be thought of as being stored in a hash
table at the entry corresponding to their Resource-ID. The peers
that make up the overlay network are also assigned an ID, called a
Peer-ID, in the same hash space as the Resource-IDs. A peer is
responsible for storing all resources that have Resource-IDs near
the peer's Peer-ID. The hash space is divided up so that all of the
hash space is always the responsibility of some particular peer,
although as peers enter and leave the system a particular peer's
area may change. Messages are exchanged between the peers in the DHT
as the peers enter and leave to preserve the structure of the DHT
and exchange stored entries. Various DHT implementations may
visualize the hash space as a grid, circle, line, or hypercube.</t>
<t>Peers keep information about the location of other peers in the
hash space and typically know about many peers nearby in the hash
space, and progressively fewer more distant peers. We refer to this
table of other peers as a Routing Table. When a peer wishes to
operate on a resource it consults the list of peers it is aware of
and contacts the peer with the Peer-ID nearest the desired
Resource-ID. If that peer does not know how to find the resource, it
either returns information about a closer peer it knows about, or
forwards the request to a closer peer. In this fashion, the request
eventually reaches the peer responsible for the resource, which then
replies to the requester.</t>
</section>
<section title="DHT Topology">
<t>Each DHT will have a somewhat different structure, but many of
the concepts are common. The DHT defines a large space of
Resource-IDs, which can be thought of as addresses. In many DHTs,
the Resource-IDs are simply 128- or 160-bit integers. Each DHT also
has a distance metric such that we can say that Resource-ID A is
closer to Resource-ID B than to Resource-ID C.</t>
<t>Each peer in the DHT is assigned a Peer-ID and is "responsible"
for the nearby space of Resource-IDs. So, for instance, if we have a
peer P, then it could also be responsible for storing data
associated with Resource-ID P+epsilon as long as no other peer P was
closer. The DHT Resource-ID space is divided so that some peer is
responsible for each Resource-ID.</t>
</section>
<section anchor="sec.routing" title="Routing">
<t>The way routing works in a DHT is specified by the specific DHT
algorithm but the basic concepts are common to most systems. Each
peer maintains connections to some other set of peers N. There need
not be anything special about the peers in N, except that the peer
has a direct connection to them: it can reach them without going
through any other peer. When it wishes to deliver a message to some
peer P, it selects some member of N, N_i that is closer to P than
itself (as a degenerate case, P may be in N). The peer sends the
message to N_i. At this point two things can happen:</t>
<t><list style="hanging">
<t hangText="Recursive Routing:">N_i repeats the same process as
P, sending the message to one of its peers N_j. This same
process repeats until the message is delivered to N.</t>
<t hangText="Iterative Routing:">N_i consults its table of
direct connections and selects a new peer N_j which is closer to
N. It responds to the original sending peer with a redirect to
N_j. The original peer then sends the message to N_j, where the
process repeats until the sending peer is redirected to N.</t>
</list></t>
<t>The advantage of iterative routing is that it consumes less
resources for the intermediate peers; they only have to send
redirect messages rather than forwarding requests and responses. The
advantage of recursive routing is that it does not require the
sending or receiving peer to have a rich set of connections to other
nodes in the overlay. Thus, iterative routing is problematic in
NATed networks because there is no way to guarantee that a peer will
be able to form a connection to whatever peer it is redirected to.
In RELOAD, iterative routing is supported using the ROUTE-QUERY
request.</t>
<t>In most DHTs, the peers in N are selected in a particular way.
One common strategy is to have them arranged exponentially further
away from yourself so that any message can be routed in a O(log(N))
steps. The details of the routing structure depend on the DHT
algorithm, however, since it defines the distance metric and the
structure of the connection table.</t>
<t>In RELOAD, messages may either be REQUESTS or RESPONSES to
REQUESTS. Requests are routed as described above. In principle,
responses could be routed independently from requests. This is
called "Asymmetric" routing because requests and responses will
generally follow different paths through the network. Asymmetric
routing makes diagnosis of errors difficult because you need to be
able to acquire debugging information at multiple locations. In the
alternative strategy, called "Symmetric" routing, as requests travel
through the network they accumulate a history of the peers they
passed through and responses are routed in the opposite direction so
that they follow the same path in reverse. RELOAD supports both
flavors of routing.</t>
<t>Symmetric routing is easier to debug. Symmetric routing is also
required when the overlay topology is changing. For example, when a
new peer is joining the overlay, asymmetric routing cannot work
because the response would not be able to reach the new peer until
it has completed the joining process. Symmetric routing solves this
situation because the response is routed from the admitting peer
through the bootstrap peer, thus relying on a path that is already
known and established. In order to implement symmetric routing,
RELOAD provides the Via List (<xref target="sec-via-list"></xref>)
feature. Asymmetric routing, however, requires no state to be stored
in the message (as a Via List) or in on-path peers.</t>
<t>[[TODO: again, this is a topic that needs WG discussion. It seems
like there are situations where symmetric is very desirable (e.g.,
startup). It's less clear that asymmetric will have a
performance/state difference that will be significant.]]</t>
<t>[[TODO: it's not clear that symmetric/asymmetric is really that
evocative a terminology. Henning suggested key-based/trace-based but
EKR doesn't like that. Other suggestions?]]</t>
</section>
<section anchor="sec-structured-data"
title="Storing and Retrieving Structured Data">
<t>The Data Storage Layer provides operations to STORE, FETCH, and
REMOVE data. Each location in the DHT is referenced by a single
integer Resource-ID. However, each location may contain data
elements corresponding to multiple kinds (e.g., certificate, SIP
registration). Similarly, there may be multiple elements of a given
kind.</t>
<figure>
<artwork><![CDATA[
+--------------------------------+
| Resource-ID |
| |
| +------------+ +------------+ |
| | Kind 1 | | Kind 2 | |
| | | | | |
| | +--------+ | | +--------+ | |
| | | Value | | | | Value | | |
| | +--------+ | | +--------+ | |
| | | | | |
| | +--------+ | | +--------+ | |
| | | Value | | | | Value | | |
| | +--------+ | | +--------+ | |
| | | +------------+ |
| | +--------+ | |
| | | Value | | |
| | +--------+ | |
| +------------+ |
+--------------------------------+
]]></artwork>
</figure>
<t>Each kind is identified by a kind-id, which is a code point
assigned by IANA. Note that a kind may be employed by multiple
usages and new usages are encouraged to use previously defined kinds
where possible. As part of the kind definition, protocol designers
may define constraints, such as limits on size, on the values which
may be stored. For many kinds, the set may be restricted to a single
value; some sets may be allowed to contain multiple identical items
while others may only have unique items. We define the following
data models in this document, though other usages can define their
own structures:</t>
<t><list style="hanging">
<t hangText="single value:">There can be at most one item in the
set and any value overwrites the previous item.</t>
<t hangText="array:">Many values can be stored and addressed by
index.</t>
<t hangText="dictionary:">The values stored are indexed by a
key. Often this key is one of the values from the certificate of
the peer sending the STORE request.</t>
</list></t>
</section>
<section anchor="sec-join-leave-maint"
title="Joining, Leaving, and Maintenance">
<t>When a new peer wishes to join the DHT, it must have a peer-id
that it is allowed to use. It uses one of the peer-ids in the
certificate it received from the enrollment server. The main steps
in joining the DHT are:</t>
<t><list style="symbols">
<t>Forming connections to some other peers.</t>
<t>Acquiring the data values this peer is responsible for
storing.</t>
<t>Informing the other peers which were previously responsible
for that data that this peer has taken over responsibility.</t>
</list></t>
<t>First, the peer ("JP," for Joining Peer) uses the bootstrap
procedures to find some (any) peer in the DHT. It then typically
contacts the peer which would have formerly been responsible for the
peer's Resource-ID (since that is where in the DHT the peer will be
joining), the Admitting Peer (AP). It copies the other peer's state,
including the data values it is now responsible for and the
identities of the peers with which the other peer has direct
connections.</t>
<t>The details of this operation depend mostly on the DHT involved,
but a typical case would be:</t>
<t><list style="numbers">
<t>JP sends a JOIN request to AP announcing its intention to
join.</t>
<t>AP sends an JOIN response.</t>
<t>AP does a sequence of STOREs to JP to give it the data it
will need.</t>
<t>AP does UPDATEs to JP and to other peers to tell it about its
own routing table. At this point, both JP and AP consider JP
responsible for some section of the DHT.</t>
<t>JP makes its own connections to the appropriate peers in the
DHT.</t>
</list></t>
<t>After this process is completed, JP is a full member of the DHT
and can process STORE/FETCH requests.</t>
</section>
</section>
<section title="Forwarding Layer">
<t>The forwarding layer is responsible for looking at message and
doing one of three things:</t>
<t><list style="symbols">
<t>Deciding the message was destined for this peer and passing the
message up to the layer above this.</t>
<t>Looking at the peer-id that represents the next peer to send
the message too and if there is an existing connection, sending
the message over the connection.</t>
<t>Requesting the DHT Routing logic to tell the forwarding layer
which peer the message needs to be forwarded to (based on the
target peer-id or resource-id), and then sending the message.</t>
</list></t>
<section anchor="direct.connect" title="Forming Direct Connections">
<t>As described in <xref target="sec.routing"></xref>, a peer
maintains a set of direct connections to other peers in the DHT.
Consider the case of a peer JP just joining the DHT. It communicates
with the admitting peer AP and gets the list of the peers in AP's
routing table. Naively, it could simply connect to the IP address
listed for each peer, but this works poorly if some of those peers
are behind a NAT or firewall. Instead, we use the CONNECT request to
establish a connection.</t>
<t>Say that peer A wishes to form a direct connection to peer B. It
gathers ICE candidates and packages them up in a CONNECT request
which it sends to B through usual DHT routing procedures. B does its
own candidate gathering and sends back a response with its
candidates. A and B then do ICE connectivity checks on the candidate
pairs. The result is a connection between A and B. At this point, A
and B can add each other to their routing tables and send messages
directly between themselves without going through other DHT
peers.</t>
<t>In general, a peer needs to maintain connections to all of the
peers near it in the DHT and to enough other peers to have efficient
routing (the details depend on the specific DHT). If a peer cannot
form a connection to some other peer, this isn't necessarily a
disaster; DHTs can route correctly even without fully connected
links. However, a peer should try to maintain the specified link set
and if it detects that it has fewer direct connections, should form
more as required.</t>
</section>
<section anchor="sec-via-list" title="Via Lists">
<t>In a general messaging system, messages need a source and a
destination and peers need to be able to send a response to the peer
that sent the request. This can be particularly tricky in overlay
networks when a new peer is joining, or the overlay network is
stabilizing and different peers have different ideas on what the
overlay topology is. A simple and reliable way to make sure that a
response can reach the node that sent the request in these
situations is to have the response traverse the reverse path of the
request.</t>
<t>The approach used here is to have each node the request traverses
add its peer-id to the "via list" in the request. Then the response
is routed by looking at the list and using it as list of peers that
the response will be routed thorough. To support this, each message
has a destination list of nodes it needs to be routed through as
well as a via list of what nodes it has traversed.</t>
<t>When a peer receives a message from the Transport Layer, it adds
the peer-id of the node it received the message from to the end of
the via list. When a peer goes to transmit a message to the
Transport Layer, it looks at the first entry on the destination
list. If the entry is this peer, it removes this entry from the list
and looks at the next entry and if the entry is not this peer, it
sends the message to the first peer on the destination list.</t>
<t>When a peer goes to send a response to a request, it can simply
copy the via list in reverse to form the destination list for the
response if it wishes to route the response along the reverse path
as the request.</t>
<t>Peers that are willing to maintain state may do list compression
for privacy reason and to reduce the message size. They do this by
taking some number of entries off the via list and replacing them
with a unique entry that this peer can later identify. Later, if the
peer sees the unique entry in a destination list, it removes the
unique entry and replaces it with the all the entries removed from
the original via list (and reverses the order of these entries).
Note that this technique will generally require storing some
per-message state on the intermediate peer, so this is a
bandwidth/per-peer state tradeoff. The exception is if the list is
not compressed but rather the peer-ids are simply encrypted.</t>
<t>The via list approach provides several features. First it allows
a response to follow the same path as the request. This is
particularly important for peers that are sending requests while
they are joining and before other peers can route to them as well as
situations where message are being exchanged to stabilize the
overlay network. It also makes it easier to diagnose and manage the
system when all peers see the response to any request they
forward.</t>
</section>
<section title="Clients">
<t>RELOAD also allows for the possibility of client nodes. A client
is a node with a peer-id which connects to an admitting peer (or
peers) like an ordinary peer but never sends a JOIN or an UPDATE. It
is therefore in the AP's connection table but not its routing table
and never is used to store any DHT data. However, because it is
reachable through the AP, it can still send and receive messages.
The client MUST still have the usual credentials. Also, because it
never sends JOINs it is never responsible for storing data.</t>
<t>Because the client may only have a connection to a single AP,
which, due to topology shifts may no longer be the responsible peer,
clients SHOULD use symmetric routing and should advertise route
lists that contain both the AP to which they are connected and
themselves. E.g., if the client has peer-id X and the AP has peer-id
Y, the client should advertise the destination list (Y, X). This
guarantees reachability.</t>
<t>Note that clients MAY also contact APs which are not in fact
responsible for the client's peer-id.</t>
</section>
</section>
<section title="Transport Layer">
<t>The transport layer sends and receives messages over TLS and DTLS.
For TLS it simply pushes the messages into the stream. For DTLS it
takes care of fragmentation issues. The reason for including TLS is
the improved performance it can offer for bulk transport of data. The
reason for including DTLS is that the percentage of the time that two
devices behind NATs can form a direct connection without a relay is
much higher for DTLS than for TLS. If all NATs were <xref
target="I-D.ietf-behave-tcp"></xref> compliant, then TLS over TCP
would be preferred.</t>
</section>
<section title="Enrollment">
<t>Before a new user can join the DHT for the first time, they must
enroll in the P2P Network for the DHT they want to join. Enrollment
will typically be done by contacting a centralized enrollment server.
Other approaches (for instance static out of band configuration) are
possible but are outside the scope of this specification. During
enrollment a new node learns about a particular overlay, sets up a
names and credentials, and discovers the bootstrap nodes. This would
typically be done when a new peer joined an overlay for the very first
time. Bootstrap is the process that happens each time a node boots and
is how the peer finds an node that can be used to join the
overlay.</t>
<t>Before a node can join an overlay, it needs to be provided with a
name for the overlay. Some examples are "example.com", "example", and
"example.local". This name is resolved via a DNS SRV lookup for
service name p2p_enroll and a protocol of tcp. If the TLD for the name
is .local, then this DNS SRV lookup is done using <xref
target="I-D.cheshire-dnsext-multicastdns"></xref> and the service name
p2p_menroll. The intention here is to support ad hoc/local overlays.
The resulting DNS lookup will provide the address of a enrollment
server. Once this server is found, HTTPS is used to retrieve a XML
file that contains the parameters for the overlay. These include
things such as: what algorithms the overlay uses, overlay parameters,
what usages are a peer on this overlay is required to support, the
type of credentials required, addresses of credentials servers, the
root certificate for the DHT, information about the DHT algorithm that
is being used, a P2P-Network-Id that uniquely identifies this ring,
and any other parameters it may need to connect to the DHT. The DHT
also informs the peers what Usages it is required to support to be a
peer on this P2P Network. It also provides an initial list of
bootstrap nodes that consists of multiple bootstrap entries that each
have the IP address and port for contacting a bootstrap peer. Some of
the address may be multicast addresses. In the case of multicast DNS,
every peer may also act as an enrollment server.</t>
<t>If shared-key security (<xref target="sec-shared-key"></xref>) is
being used, then the peer can proceed directly to bootstrap. If
certificate-based security (<xref target="sec-cert-security"></xref>
is being used, the peer MUST contact the credential server to obtain a
certificate.</t>
<section title="Certificate Issuance">
<t>Once the peer has the XML file that identifies if credentials are
needed, it can contact the credential server. The user establishes
his identity to the server's satisfaction and provides the server
with its public key. The centralized server then returns a
certificate binding the user's user name to his public key. The
properties of the certificate are discussed in <xref
target="sec-security-intro"></xref>. The amount of authentication
performed here can vary radically depending on the DHT network being
joined. Some networks may do no verification at all and some may
require extensive identity verification (e.g., checking a driver's
license) before issuing credentials for a given user name. The only
invariant that the enrollment server needs to ensure is that no two
users may have the same identity.</t>
</section>
<section title="Bootstrap">
<t>The above steps are only done the first time a peer joins a new
overlay or when the overlay parameters are close to their expiration
time (as listed in the configuration document) and need to be
refreshed. The next step is the bootstrap step which is done every
time the peer boots.</t>
<t>Bootstrapping consists of looking at the list of cached nodes and
bootstraps nodes and sending a RELOAD PING to them to see if they
respond. Once a node responds, it can be used to join the overlay.
After a node has joined, it keeps track of a small number of peers
to which it could directly connect. Theses are saved as the cached
nodes and used next time the peer boots. The point of the cached
nodes is to reduce the load on the bootstrap nodes.</t>
</section>
</section>
<section anchor="sec-security-intro" title="Security">
<section anchor="sec-cert-security" title="Certificate-Based Security">
<t>The certificate-based security model revolves around the
enrollment process allocating a unique name to the user and issuing
a certificate <xref target="RFC3280"></xref> for a public/private
key pair for the user. All peers in a particular DHT can verify
these certificates. A given peer acts on behalf of a user, and that
user is responsible for its operation.</t>
<t>The certificate serves two purposes:</t>
<t><list style="symbols">
<t>It entitles the user to store data at specific locations in
the DHT. Each usage defines the specific rules for determining
which certificates can access each Hashed-ID/kind-id pair. For
instance, some usages might allow anyone to write at a given
location, whereas others might restrict writes to a single
certificate.</t>
<t>It entitles the user to operate a peer that has a peer-id
found in the certificate. When the peer is acting as a DTLS or
TLS server, it can use this certificate so that a client
connecting to it knows it is connected to the correct
server.</t>
</list></t>
<t>When a user enrolls, or enrolls a device with no keying material,
the user is given a certificate. This certificate contains
information that identifies the user and the device they are using.
If a user has more than one device, typically they would get one
certificate for each device. This allows each device to act as a
separate peer.</t>
<t>The contents of the certificate include:</t>
<t><list style="symbols">
<t>A public key provided by the user.</t>
<t>Zero or more user names that the DHT is allowing this user to
use. For example, "alice@example.org". Typically a certificate
will have one name. In the SIP usage, this name corresponds to
the AOR.</t>
<t>Zero or more peer-ids. Typically there will be one peer-id.
Each device will use a different peer-id, even if two devices
belong to the same user. Peer-IDs should be chosen randomly by
the enrollment server.</t>
<t>A serial number that is unique to this certificate across all
the certificates issued for this DHT.</t>
<t>An expiration time for the certificate.</t>
</list></t>
<t>Note that because peer-IDs are chosen randomly, they will be
randomly distributed with respect to the user name. This has the
result that any given peer is highly unlikely to be responsible for
storing data corresponding to its own user, which promotes high
availability.</t>
<section title="Storage Permissions">
<t>When a peer uses a STORE request to place data at a particular
location X, it must sign with the private key that corresponds to
a certificate that is suitable for storing at location X. Each
data kind in a usage defines the exact rules for determining what
certificate is appropriate.</t>
<t>The most natural rule is that a certificate with user name X
"owns" data located at Hash(X) (X is the Unhashed-ID and Hash(X)
is the Hashed-ID) and only he can write there. This rules is used
for all the kinds defined in this specification. Thus, only a user
with a certificate for "alice@example.org" could write to that
location in the DHT. However, other usages can define any rules
they choose, including publicly writable values.</t>
<t>The digital signature over the data serves two purposes. First,
it allows the peer responsible for storing the data to verify that
this STORE is authorized. Second, it provides integrity for the
data. The signature is saved along with the data value (or values)
so that any reader can verify the integrity of the data. Of
course, the responsible peer can "lose" the value but it cannot
undetectably modify it.</t>
</section>
<section title="Peer Permissions">
<t>The second purpose of a certificate is to allow the device to
act as a peer with the specified peer-ID. When a peer wishes to
connect to peer X, it forms a TLS/DTLS connection to the peer and
then performs TLS mutual authentication and verifies that the
presented certificate contains peer-ID X.</t>
<t>Note that because the formation of a connection between two
nodes generally requires traversing other nodes in the DHT, as
specified in <xref target="direct.connect"></xref>, those nodes
can interfere with connection initiation. However, if they attempt
to impersonate the target peer they will be unable to complete the
TLS mutual authentication: therefore such attacks can be
detected.</t>
</section>
<section title="Expiry and Renewal">
<t>At some point before the certificate expires, the user will
need to get a new certificate from the enrollment server.</t>
</section>
</section>
<section anchor="sec-shared-key" title="Shared-Key Security">
<t>RELOAD also defines a shared-key security model which can be used
in closed networks where the peers are not mutually suspicious. In
this model, the peers all share a single key which is used to
authenticate the peer-to-peer DTLS connections via TLS-PSK/TLS-SRP.
If shared-key security mode is in use, a shared-key capable cipher
suite such as TLS-PSK or TLS-SRP MUST be used. This is useful for
admission control, but is completely unsafe in any setting where
peers are not mutually trusted, since it allows any peer to
impersonate any other peer.</t>
</section>
</section>
<section title="Migration">
<t>At some point in time, a given P2P Network may want to migrate from
one underlying DHT algorithm to another or update to a later extension
of the protocol. This can also be used for crypto agility issues. The
migration approach is done by having peers initializing algorithm A.
When the clients go to periodically renew their credentials, they find
out that the P2P Network now requires them to use algorithm A but also
to store all the data with algorithm B. At this point there are
effectively two DHT rings in use, rings A and B. All data is written
to both but queries only go to A. At some point when the clients
periodically renew their credentials, they learn that the P2P Network
has moved to storing to both A and B but that FETCH requests are done
with P2P Network B and that any SEND should first be attempted on P2P
Network B and if that fails, retried on P2P Network A. In the final
stage when clients renew credentials, they find out that P2P Network A
is no longer required and only P2P Network B is in use. Some types of
usages and environments may be able to migrate very quickly and do all
of these steps in under a week, depending on how quickly software that
supports both A and B is deployed and how often credentials are
renewed. On the other hand, some very ad-hoc environments involving
software from many different providers may take years to migrate.</t>
<t>[[TODO: This needs more filling out]]</t>
</section>
<section title="Usages Layer">
<t>By itself, the distributed storage layer just provides
infrastructure on which applications are built. In order to do
anything useful, a usage must be defined. Each Usage needs to specify
several things:</t>
<t><list style="symbols">
<t>Register kind-id code points for any kinds that the Usage
defines.</t>
<t>Define the data structure for each of the kinds.</t>
<t>Define access control rules for each kinds.</t>
<t>Provide a size limit for each kinds.</t>
<t>Define how the Unhashed-ID is formed that is hashed to form the
Resource-ID where each kind is stored.</t>
<t>Describe how values will be merged after a network partition.
Unless otherwise specified, the default merging rule is to act as
if all the values that need to be merged were stored and that the
order they were stored in corresponds to the stored time values
associated with (and carried in) their values. Because the stored
time values are those associated with the peer which did the
writing, clock skew is generally not an issue. If if two nodes are
on different partitions, clocks, this can create merge conflicts.
However because RELOAD deliberately segregates storage so that
data from different users and peers is stored in different
locations, and a single peer will typically only be in a single
network partition, this case will generally not arise.</t>
</list></t>
<t>The kinds defined by a usage may also be applied to other usages.
However, a need for different parameters, such as different size
limits, would imply the need to create a new kind.</t>
<section anchor="sec-sip-usage" title="SIP Usage">
<t>From the perspective of P2PSIP, the most important usage is the
SIP Usage. The basic function of the SIP usage is to allow Alice to
start with a SIP URI (e.g., "bob@dht.example.com") and end up with a
connection which Bob's SIP UA can use to pass SIP messages back and
forth to Alice's SIP UA.</t>
<!--
<t>This operation can take a number of forms, but in the simplest
case, Bob's SIP UA has peer-ID "B". When Bob joins the DHT (i.e.,
turns on his phone), he stores the following mapping in the DHT:</t>
<t><list style="symbols">
<t>sip:bob@dht.example.com -> B</t>
</list></t>
<t>When Alice wants to call Bob, she starts with his URI and her UA
uses the DHT to look up his peer-ID B. She then routes a message
through the DHT to B requesting a direct connection. Once this
connection is established she can send SIP messages over it, which
allows her to set up the phone call.</t>
-->
<t>This is done using three key operations that are provided by the
SIP Usage. They are:</t>
<t><list style="symbols">
<t>Mapping SIP URIs that are not GRUUs to other SIP URIs or to
the DHT peer responsible for the SIP UA.</t>
<t>Mapping SIP GRUUs to the DHT peer responsible for the SIP
UA.</t>
<t>Forming a connection directly to a DHT peer that is used to
send SIP messages to the SIP UA.</t>
</list></t>
<t>All SIP URIs for a given overlay MUST be constructed so that they
terminate in the domain name of the overlay. For instance, if the
overlay name is "example.com", then all AORs must be of the form
{sip,sips}:username@example.com. Accordingly, to dereference a URI,
a P2PSIP implementation MUST check to see if the domain matches an
overlay which it is a member of. If so, it uses the following
procedures. Otherwise, it MUST follow <xref target="RFC3263"></xref>
procedures. Note that unless the P2PSIP overlay provides some kind
of gateway to ordinary SIP (e.g., a publicly accessible SIP server)
this is likely to be only partially successful, since, for instance,
the callee may not be able to call back.</t>
<section title="SIP Location">
<t>A peer acting as a SIP UA stores their registration information
in the DHT by storing either another URI (for retargeting) or a
destination lists to reach them at a Resource-ID in the DHT formed
from the user's SIP AOR. When another peer wishes to find a peer
that is registered for a SIP URI, the lookup of the user's name is
done by taking the user's SIP Address or Record (AOR) and using it
as the Unhashed-ID that is hashed to get a Resource-ID. When the
Unhashed-ID is dereferenced, the result is a set of values. Each
value is either another SIP URI or a destination list. If the
value is a SIP URI, the calling peer looks up that URI and
continues the process until it gets a destination list.</t>
<t>If the value is a destination list, then it is used to reach a
peer that represents a SIP UA registered for that AOR. Typically
this destination list will have just one entry but in the case of
peers or clients that can not be directly reached (for instance
via a strict NAT or firewall), a destination list with more than
one entry may need to be used.</t>
<t>The Unhashed-ID for this usage is a user's SIP AOR, such as
"sip:alice@example.com". This allows the set to store many values
in a dictionary structure. The authorization policy is that STORE
requests are only allowed if the user name in the signing
certificate, when turned into a SIP URL and hashed, matches the
Resource-ID. This policy ensures that only a user with the
certificate with the user name "alice@example.com" can write to
the Resource-ID that will be used to look up calls to
"sip:alice@example.com".</t>
<t>[[Open Issue: Should the Unhashed-ID be
"sip:alice@example.com", "alice@example.com", or a string that
includes the code point defined for the kind? The issue here is
determining whether different usages that store data at a
Unhashed-ID that is primarily formed from "alice@example.com"
should hash to the same Resource-ID as the SIP Usage. For example,
if a buddy list had a Unhashed-ID that was roughly the same, would
we want the buddy list information to end up on the same peers
that stored the SIP location data or on different peers?]]</t>
</section>
<section title="SIP GRUUs">
<t>GRUUs that refer to peers in the P2P network are constructed by
simply forming a GRUU, where the value of gr URI parameter
contains a base64 encoded version of the destination list that
will reach the peer. Typically the destination list is just a
single entry with the peer-id of peer.</t>
</section>
<section title="SIP Connect">
<t>This usage allows two clients to form a new TLS or DTLS
connection between them and then use this connection for sending
SIP messages to one another. This does not store any information
in the DHT, but it allows the CONNECT request to be used to set up
a TLS or DTLS connection between two peers and then use that
connection to send SIP messages back and forth.</t>
<t>The CONNECT request will ensure that the connection is formed
to a peer that has a certificate which includes the user that the
connection is being formed to.</t>
</section>
<section title="SIP Tunnel">
<!-- EKRTODO: remove?-->
<t>This TUNNEL request allows two peers to exchange SIP messages
across the overlay using the TUNNEL method without first setting
up a direct connection using CONNECT. This allows a SIP message to
be sent immediately, without the delay associated with CONNECT and
for a simple SIP exchange, it may result in fewer messages being
sent.</t>
</section>
</section>
<section title="Certificate Store Usage">
<t>This usage allows each user to store their certificate in the DHT
so that it can be retrieved to be checked by various peers and
applications. Peers acting on behalf of a particular user store that
user's certificate in the DHT, and any peer that needs the
certificate can do a FETCH to retrieve the certificate. Typically it
is retrieved to check a signature on a request or the signature on a
chunk of data that the DHT has received.</t>
</section>
<section title="TURN Usage">
<t>This usage defines a new kind for finding STUN-Relay servers. Any
peer that supports this usage saves a pointer to the IP address and
port of the TURN server in the DHT. When a peer wishes to discover a
TURN server, it picks a random Resource-ID and performs a FIND at
that Resource-ID for the appropriate type for the service. If
nothing is found, this can be repeated until an appropriate set of
servers are found.</t>
</section>
<section title="Diagnostic Usage">
<t>This usage defines several new kinds that be queried to find
information about the peer that may be useful for monitoring and
diagnostics. This includes information such as software version,
neighbor information, and performance statistics.</t>
</section>
<section title="HIP Tunnel Usage">
<t>This usage allows two peers running HIP to tunnel HIP messages
across the overlay. This allows the HIP peers to use the overlay as
a rendezvous system to set up a direct path using HIP NAT traversal
mechanisms.</t>
</section>
<figure>
<!-- begin-prologue -->
<artwork><![CDATA[
primitive uint8 8;
primitive uint16 16;
primitive uint24 24;
primitive uint32 32;
primitive int32 32;
primitive uint64 64;
primitive uint128 128;
primitive char 8;
primitive opaque 8;
primitive blob 0;
typedef char string<65000>;
primitive peer_id 128;
typedef opaque resource_id<255>;
typedef uint32 overlay;
typedef uint64 transaction_id;
typedef uint32 kind_id;
typedef uint64 generation_counter;
struct {
uint32 addr;
uint16 port;
} ip4_addr_port;
struct {
uint128 addr;
uint16 port;
} ip6_addr_port;
enum {ip4_address_type (1), ip6_address_type (2)} address_type;
select {
case ip4_address_type:
ip4_addr_port v4;
case ip6_address_type:
ip6_addr_port v6;
} ip_address_and_port;
]]></artwork>
<!-- end-prologue -->
</figure>
</section>
</section>
<section anchor="sec-base-protocol" title="Base Protocol">
<t>RELOAD is a message-oriented request/response protocol. The messages
are encoded using binary fields. All integers are represented in network
byte order. The general philosophy behind the design was to use Type,
Length, Value fields to allow for extensibility. However, for the parts
of a structure that were required in all messages, just define theses in
a fixed position as adding a type and length for them is unnecessary and
would simply increase bandwidth and introduces new potentials for
interoperability issues.</t>
<t>Each message has three parts:</t>
<t><list style="hanging">
<t hangText="Forwarding Header:">Each message has a generic header
which is used to forward the message between peers and to its final
destination. This header is the only information that an
intermediate peer (i.e., one that is not the target of a message)
needs to examine.</t>
<t hangText="Message Contents:">The message being delivered between
the peers. From the perspective of the forwarding layer, the
contents is opaque, however, it is interpreted by the higher
layers.</t>
<t hangText="Signature:">A digital signature over the message
contents and parts of the header of the message. Note that this
signature can be computed without parsing the message contents.</t>
</list></t>
<t>The following sections describe the format of each part of the
message.</t>
<section anchor="sec-forwarding-header" title="Forwarding Header">
<t>The layout of the forwarding header is shown below</t>
<figure>
<artwork><![CDATA[
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|1| R | E | L | O |
4 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Overlay |
8 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| | |F|L| |
| TTL | Reserved |R|F| Fragment Offset |
| | |A|R| |
| | |G|G| |
12 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| | |
| Version | Length |
| | |
16 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Transaction ID |
+ +
| |
24 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Via | Route | |
| List | List | Flags |
| Length | Length | |
28 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| |
// Via List //
| |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| |
// Destination List //
| |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| |
// Route Log //
| |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
]]></artwork>
</figure>
<t>The first four bytes identify this message as a RELOAD message. The
message is easy to demultiplex from STUN messages by looking at the
first bit.</t>
<t>The Overlay field is the 32 bit checksum/hash of the overlay being
used. The variable length string representing the overlay name is
hashed with SHA-1 and the low order 32 bits are used. The purpose of
this field is to allow nodes to participate in multiple overlays and
to detect accidental misconfiguration.</t>
<t>TTL (time-to-live) is an 8 bit field indicating the number of
iterations, or hops, a message can experience before it is discarded.
The TTL value MUST be decremented by one at every hop along the route
the message traverses. If the TTL is 0, the message MUST NOT be
propagated further and MUST be discarded. The initial value of the TTL
should be TBD.</t>
<t>FRAG is a 1 bit field used to specify if this message is a
fragment.</t>
<figure>
<artwork><![CDATA[
NOT-FRAGMENT : 0x0
FRAGMENT : 0x1
]]></artwork>
</figure>
<t>LFRG is a 1 bit field used to specify whether this is the last
fragment in a complete message.</t>
<figure>
<artwork><![CDATA[
NOT-LAST-FRAGMENT : 0x0
LAST-FRAGMENT : 0x1
]]></artwork>
</figure>
<t>[[Open Issue: How should the fragment offset and total length be
encoded in the header? Right now we have 14 bits reserved with the
intention that they be used for fragmenting, though additional bytes
in the header might be needed for fragmentation.]]</t>
<t>Version is a 7 bit field that indicates the version of the RELOAD
protocol being used.</t>
<figure>
<artwork><![CDATA[
Version1.0 : 0x1
]]></artwork>
</figure>
<t>The message Length is the count in bytes of the size of the
message, including the header.</t>
<t>The Transaction ID is a unique 64 bit number that identifies this
transaction and also serves as a salt to randomize the request and the
response. Responses use the same Transaction ID as the request they
correspond to. Transaction IDs are also used for fragment
reassembly.</t>
<t>The Destination List Length and the Via List Length contain the
lengths of the route and via lists respectively, in the number of
objects.</t>
<t>[[Open Issue: How should we handle peer-id lengths? This basically
assumes they're fixed length per DHT algorithm (but not fixed-length
for RELOAD) so that you can unambiguously parse things. Should we have
a length byte?]]</t>
<t>The flags word contains control flags. There is one currently
defined flag.</t>
<figure>
<artwork><![CDATA[
ROUTE-LOG : 0x1
]]></artwork>
</figure>
<t>The ROUTE-LOG flag indicates that the route log should be included
(see <xref target="sec-route-log"></xref></t>
<t>The Destination List contains a sequence of destinations which the
message should pass through. The destination list is constructed by
the message originator. The first element in the destination list is
where the message goes next. The list shrinks as the message traverses
each listed peer. Destinations are defined at the end of this
section.</t>
<t>The Via List contains the sequence of destinations through which
the message has passed. The via list starts out empty and grows as the
message traverses each peer.</t>
<t>If a message was being sent thought the sequences of peers A,B,C,D,
the message from A to B would have a empty via list and a route of
list of B,C,D. The message from B to C would have a via list of A then
route of C,D and so on. This means that when the route list is
followed exactly, all that is needed to update these lists is to
change their lengths. This avoids the need to change or move any of
the other list entries. In other cases, some entries may need to be
copied or moved.</t>
<t>The destination list and via lists a list of objects of type
destination_object:</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
enum {peer(0), compressed(1), resource(2)} destination_type;
select {
case peer:
peer_id peer_id;
case compressed:
opaque compress<255>;
case resource:
opaque resource_id<255>;
} dest_object;
public struct {
dest_object info;
} destination_object;
]]></artwork>
</figure>
<t>A destination_object can have one of three types (this is
extensible):</t>
<t><list style="hanging">
<t hangText="peer:">A peer-id. All peer-ids are of fixed length
for a given overlay and therefore peer-ids have no length.</t>
<t hangText="compressed:">A compressed list of peer-ids and/or
resources. This value is variable length but because it was
compressed by one of the peers, it is only meaningful to that peer
and cannot be decoded by other peers.</t>
<t hangText="resource:">The resource id of the resource which is
desired. This type MUST only appear in the final location of a
destination list and MUST NOT appear in a via list. It is
meaningless to try to route through a resource.</t>
</list></t>
<section anchor="sec-reload-diffs"
title="Changes to Forwarding Header">
<t>The RELOAD-01 forwarding header was completely fixed, whereas
this header includes lists that change en-route. However, this type
of operation is easily accomplished in both software and hardware,
therefore we still view it as a low-overhead header. The changes
include the following.</t>
<t><list style="symbols">
<t>Rearranged fields to have a cleaner separation between
payload and header.</t>
<t>Removed DHT, Hash, and Security parameters. These are now in
the overlay bootstrap system rather than per-message.</t>
<t>Source and destination IDs are now destination lists to
accommodate source routing and recursion without state on
intermediate peers.</t>
<t>Added route log to header to allow payload/header
separation.</t>
</list></t>
</section>
<section anchor="sec-transport-routing" title="Message Routing">
<section anchor="sec-request-origination"
title="Request Origination">
<t>In order to send a message to a given peer-id or resource-id, a
peer must construct an appropriate destination list. The most
common such destination list is a single entry containing the
peer/resource-id. This simply uses the normal DHT routing
mechanisms to forward the message to that destination.</t>
<t>Messages can also be source routed. In order to construct a
source route, the originator provides a destination list
containing a sequence of resource-ids. The semantics of this
destination list are that the message is to traverse in order
(potentially with intermediate hops) each entry on the destination
list. As each peer is traversed, that entry is removed from the
destination list. This makes it possible to address a peer which
is potentially behind a NAT or a firewall in such a way that it
cannot be connected to directly under any circumstances.</t>
<t>[[TODO: Salman has suggested the originator doing parallel
requests/responses. This is an open issue.]</t>
</section>
<section anchor="sec-response-origination"
title="Response Origination">
<t>When a peer sends a response to a request, it SHOULD construct
the destination list by reversing the order of the entries on the
via list. This has the result that the response traverses (at
least) the same peers as the request traversed, except in reverse
order (symmetric routing). For asymmetric routing, the peer MAY
simply use the first entry on the via list.</t>
</section>
<section anchor="sec-message-forwarding"
title="Message Receipt and Forwarding">
<t>When a peer receives a message, it first examines the overlay,
version, and other header fields to determine whether the message
is one it can process. If any of these are incorrect (e.g., the
message is for an overlay in which the peer does not participate)
it is an error. The peer SHOULD generate an appropriate error but
MAY simply drop the message.</t>
<t>Once the peer has determined that the message is correctly
formatted, it examines the first entry on the destination list.
There are three possible cases here:</t>
<t><list style="symbols">
<t>The first entry on the destination list is a private id
which is being used for destination list compression.</t>
<t>The first entry on the destination list is an id for which
the peer is responsible.</t>
<t>The first entry on the destination list is for which
another peer is responsible.</t>
</list></t>
<t>These cases are handled separately.</t>
<section anchor="sec-private-peer-id" title="Private ID">
<t>If the first entry on the destination list is a private id,
the peer replaces that entry with the store local value that it
indexes and then re-examines the destination list to determine
which case now applies.</t>
</section>
<section anchor="sec-responsible-id" title="Responsible ID">
<t>If the first entry on the destination list is a Hashed-ID for
which the peer is responsible, the peer strips the entry off the
route list. If there are remaining entries on the destination
list, the peer then re-examines the destination list to
determine which case now applies. If the destination list is now
empty, then the message was destined for this peer and it MUST
pass it to the next layer up.</t>
</section>
<section anchor="sec-other-id" title="Other Hashed-ID">
<t>If neither of the other two cases applies, then the peer MUST
forward the message towards the first entry on the destination
list. This means that it MUST select one of the peers in its
route table which is closer to the first entry than to itself
and send the message to that peer. If the first entry on the
destination list is in the peer's connection table, then it
SHOULD forward the message to that peer directly.</t>
<t>When forwarding a message, the peer MUST:</t>
<t><list style="symbols">
<t>Decrement the TTL value</t>
<t>Update the via list.</t>
</list></t>
<t>The natural way to update the via list is simply to add the
peer-id of the peer from which the message was received to the
end of the list. However, peers may use any algorithm of their
choice provided that if the peer received a destination list
constructed by reversing the via list it would be able to route
the outgoing message correctly, enabling symmetric routing.</t>
<t>For instance, if node D receives a message from node C with
via list (A, B), the simple approach is simply to forward to the
next node (E) with via list (A, B, C). Now, if E wants to
respond to the message, it reverses the via list to produce the
destination list, resulting in (D, C, B, A). When D forwards the
response to C, the destination list will contain (B, A).
However, node D could also list compression and send E the via
list (X). E would then use the destination list (D, X). When D
processes this destination list, it MUST detect that X is a
compressed entry, recover the via list (A, B, C), and reverse
that to produce the correct destination list (C, B, A) before
sending it to C.</t>
<t>Note that if a peer is using list compression and then exits
the overlay, the message cannot be forwarded and will be
dropped. The ordinary timeout and retransmission networks
provide stability over this type of failure.</t>
</section>
</section>
</section>
<section anchor="sec-frag-reass" title="Fragmentation and Reassembly">
<t>In order to allow transport over datagram protocols, RELOAD
messages may be fragmented. If a message is too large for a peer to
transmit to the next peer it MUST fragment the message. Note that
this implies that intermediate peers may re-fragment messages if the
incoming and outgoing paths have different maximum datagram sizes.
Intermediate peers SHOULD NOT reassemble fragments.</t>
<t>Upon receipt of a fragmented message by the intended peer, the
peer holds the fragments in a holding buffer until the entire
message has been received. The message is then reassembled into a
single unfragmented message and processed. In order to prevent
denial of service attacks, receivers SHOULD time out incomplete
fragments. [[TODO: Describe algorithm]]</t>
</section>
<section anchor="sec-route-log" title="Route Logging">
<t>The route logging feature provides diagnostic information about
the path taken by the request so far and in this manner it is
similar in function to <xref target="RFC3261">SIP's</xref> Via
header field. If the ROUTE-LOG flag is set in the Flags word, at
each hop peers MUST append a route log entry to the route log
element in the header. The order of the route log entry elements in
the message is determined by the order of the peers were traversed
along the path. The first route log entry corresponds to the peer at
the first hop along the path, and each subsequent entry corresponds
to the peer at the next hop along the path. If the ROUTE-LOG flag is
set in a request, the route log MUST be copied into the response and
the ROUTE-LOG flag set so that the originator receives the ROUTE-LOG
data.</t>
<t>If the responder wishes to have a route log in the reverse
direction, it MAY set the ROUTE-LOG flag in its response as well.
Note, however, that this means that the response will grow on the
return path, which may potentially mean that it gets dropped due to
becoming too large for some intermediate hop. Thus, this option must
be used with care.</t>
<t></t>
<figure>
<!-- begin-pdu-->
<artwork><![CDATA[
public struct {
route_log_entry entries<65000>;
} route_log;
]]></artwork>
</figure>
<t>The route log is simply a variable length list of route log
entries. The first two bytes are the length, followed by a sequence
of route leg entries, each of which may be individually parsed.</t>
<figure>
<!-- begin-pdu-->
<artwork><![CDATA[
struct {
peer_id id;
uint32 uptime;
opaque certificate<65000>;
ip_address_and_port address;
} peer_info_data;
public struct {
string version;
uint8 transport;
peer_info_data peer_info;
} route_log_entry;
]]></artwork>
</figure>
<t>Each route log entry consists of the following values:</t>
<t><list style="hanging">
<t hangText="Version -">A textual representation of the software
version</t>
<t hangText="Transport -">The transport type, 1 for TLS, 2 for
DTLS</t>
<t hangText="Id -">The peer-id of the peer.</t>
<t hangText="Uptime -">The uptime of the peer in seconds.</t>
<t hangText="Certificate -">The peer's certificate. Note that
this may be omitted by setting the length to zero.</t>
<t hangText="Address -">The address and port of the peer. This
can be either an IPv4 or IPv6 address.</t>
</list></t>
<!--
route-log = uint16 ; number of route-log-entries
*route-log-entry
route-log-entry = route-log-p2psip-version
route-log-entry-transport
route-log-entry-peer-info
route-log-version = string
route-log-entry-transport = route-log-transport-tcp /
route-log-transport-udp /
route-log-transport-tls /
route-log-transport-dtls / uint8
route-log-transport-tcp = 1
route-log-transport-udp = 2
route-log-transport-tls = 3
route-log-transport-dtls = 4
route-log-entry-peer-info = peer-id
peer-name
peer-uptime
peer-ip-port
peer-certificate
peer-name = string ; a textual representation of the peer name
peer-ip-port = string ; a textual representation of the IP/port
peer-uptime = uint32 ; time in seconds since peer start
peer-certificate = data-value ; the peer's certificate -->
</section>
</section>
<section anchor="sec-contents" title="Message Contents Format">
<t>Although from the perspective of the forwarding layer the content
is opaque, all RELOAD messages share a common content structure
consisting of two parts:</t>
<t><list style="hanging">
<t hangText="Common Header:">A common header containing the
request method/response code, and a transaction ID.</t>
<t hangText="Payload:">The actual body of the request/response.
These are dependent on whether this is a request or response and
the type of request being carried.</t>
</list></t>
<section anchor="sec-contents-common-header" title="Common Header">
<t>The layout of the common header is shown below:</t>
<figure>
<artwork><![CDATA[
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| | |
| Message Code | Reserved |
| | |
4 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
]]></artwork>
</figure>
<t>Message Code is a 16 bit field that indicates which message this
is. This field is broken up as follows:</t>
<t><list style="hanging">
<t hangText="0">Reserved</t>
<t hangText="1 .. 0x7fff">Requests and responses. These code
points are always paired, with requests being odd and the
corresponding response being the request code plus 1. Thus,
PING_Q (the PING request) has value 1 and PING_A (the PING
response) has value 2</t>
<t hangText="0x8000 .. 0xfffe">Reserved</t>
<t hangText="0xffff">Error</t>
</list></t>
</section>
<section anchor="sec-payload" title="Payload">
<t>Payload is a simple string of uninterpreted bytes preceded by a
length field indicating the length of the data, not including the
length field. The bytes themselves are dependent on the code
value.</t>
<figure>
<artwork><![CDATA[
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| |
| Length |
| |
4 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| |
// Length bytes of data //
| |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
]]></artwork>
</figure>
</section>
<section anchor="sec-signature" title="Signature">
<t>The Signature element is used to attach signatures to messages
and or stored data elements. All signatures are formatted using this
element. However, the input structure to the signature computation
varies depending on the data element being signed.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
uint8 algorithm;
opaque signature_value<65000>;
signature_identity identity;
} signature;
]]></artwork>
</figure>
<t>The signature construct is just a container for the signature. It
contains the following values:</t>
<t><list style="hanging">
<t hangText="Algorithm -">The signature algorithm in use. This
may have the values RSA-SHA1 (0x01) or RSA-SHA-256 (0x02).</t>
<t hangText="Value - ">The signature value itself. This is just
the string of bytes emitted by the signature algorithm.</t>
<t hangText="Identity -">The identity or certificate used to
form the signature</t>
</list></t>
<t>[[TODO: Should we convert all of this to CMS?]]</t>
<t>A number of possible identity formats are permitted, as shown
below. The peer may indicate any of:</t>
<t><list style="symbols">
<t>Peer-id</t>
<t>User name</t>
<t>The certificate itself.</t>
</list></t>
<t>The first byte of the identity field is a type indicating the
type of identity in use.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
enum {signer_identity_peer (1), signer_identity_name (2),
signer_identity_certificate (3)} signer_identity_type;
select {
case signer_identity_peer:
peer_id id;
case signer_identity_name:
string signer_name;
case signer_identity_certificate:
opaque certificate<65000>;
} signer_identity;
public struct {
signer_identity signer;
} signer_object;
]]></artwork>
</figure>
<!--
signature = signature-algorithm
signer-identity
signature-value
signature-algorithm = rsa-sha1-id / rsa-sha-256-id / uint16
rsa-sha1-id = 0x01
rsa-sha256-id = 0x02
signer-identity = signer-identity-type
peer-id / signer-name / certificate / *byte
signer-identity-type = signer-identity-peer / signer-identity-name /
signer-identity-certificate / uint8
signer-identity-peer = 0x01
signer-identity-name = 0x02
signer-identity-certificate = 0x03
signer-name = string
certificate = data-value
signature = data-value
-->
<t>For signatures over messages the input to the signature function
is:</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
overlay overlay;
transaction_id xid;
blob signer_identity;
blob message_contents;
} message_signature_input;
]]></artwork>
</figure>
<t>The contents of this structure are as follows:</t>
<t><list style="hanging">
<t hangText="Overlay -">The overlay identifier from the
message.</t>
<t hangText="Xid -">The transaction id from the message.</t>
<t hangText="Signer Identity -">The identify of the signer (from
the signature structure.)</t>
<t hangText="Message Contents -">The contents section of the
message.</t>
</list></t>
<!--
message-signature-input = overlay ; from forwarding header
transaction-id ; from forwarding header
signer-identity
message-contents
-->
<t>[[TODO: Check the inputs to this carefully.]]</t>
<t>The input to signatures over data values is different, and is
described in <xref target="sec-data-sig"></xref>.</t>
</section>
</section>
<section anchor="sec-response-code"
title="Response Codes and Response Errors">
<t>A peer processing a request returns its status in the Message Code
field of the common header. If the request was a success, then the
message code is the response code that matches the request (i.e., the
next code up). The response payload is then as defined in the
request/response descriptions.</t>
<t>If the request failed, then the message code is set to 0xffff
(error) and the payload MUST be an error_response PDU, as shown
below.</t>
<t>For any code other than 200, the payload should be as defined
below:</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
uint16 error_code;
string reason_phrase;
opaque error_info<65000>;
} error_response;
]]></artwork>
</figure>
<!--
error-payload = string ; reason phrase
error-payload-rest
error-payload-rest = error-payload-302 / *byte
error-payload-302 = peer-id -->
<t>The contents of this payload are:</t>
<t><list style="hanging">
<t hangText="Error Code - ">A numeric error code indicating the
error that occurred"</t>
<t hangText="Reason Phrase -">A free form text string indicating
the reason for the response. The reason phrase SHOULD BE as
indicated in the error code list (e.g., "Moved Temporarily).</t>
<t hangText="Error Info -">Payload specific error information.
This MUST be empty except as specified below.</t>
</list></t>
<t>The following error code values are defined. [[TODO: These are
currently semi-aligned with SIP codes. that's probably bad and we need
to fix.]</t>
<t><list style="hanging">
<t hangText="302 (Moved Temporarily):">The requesting peer SHOULD
retry the request at the new address specified in the 302 response
message.</t>
<t hangText="401 (Unauthorized):">The requesting peer needs to
sign and provide a certificate. [[TODO: The semantics here don't
seem quite right.]]</t>
<t hangText="403 (Forbidden):">The requesting peer does not have
permission to make this request.</t>
<t hangText="404 (Not Found):">The resource or peer cannot be
found or does not exist.</t>
<t hangText="408 (Request Timeout):">A response to the request has
not been received in a suitable amount of time. The requesting
peer MAY resend the request at a later time.</t>
<t hangText="412 (Precondition Failed):">A request can't be
completed because some precondition was incorrect. For instance,
the wrong generation counter was provided</t>
<t hangText="498 (Incompatible with Overlay)">A peer receiving the
request is using a different overlay, DHT algorithm, or hash
algorithm. [[Open Issue: What is the best error number and reason
phrase to use?]]</t>
<t hangText="499 (UnWilling To Proxy)">A peer receiving the
request is unwilling to support the Routing mechanism specified in
the Routing field of the message header. [[Open Issue: What is the
best error number and reason phrase to use?]]</t>
</list></t>
</section>
</section>
<section title="End-to-End Timeout and Retransmission">
<t>Timeout and retransmission are handled on an end-to-end basis as well
as the transports providing a hop by hop reliability mechanism. For
end-to-end reliability, the requesting node retransmits a requests every
3 seconds until it receives a response or after it has send the request
5 times. Retransmissions MUST use the same transaction ID.</t>
</section>
<section anchor="sec-transports" title="Transports">
<t>Currently multiple transport protocols are specified and more may be
defined in the future. Implementation MUST implement TLS and DTLS. A
given overlay can choose which protocols it uses.</t>
<section title="TLS">
<t>TLS runs on top of TCP which offers the best performance from a
data transfer point of view and does not require as frequent keep
alive messages.</t>
</section>
<section title="DTLS">
<t>DTLS runs on top of UDP which offers the highest probability of
direct connectivity in the face of the current generation of consumer
NATs.</t>
<section title="Reliability for Unreliable Transports">
<t>When RELOAD is carried over DTLS or another unreliable transport,
it needs to be used with a reliability and flow control mechanism,
which is provided on a hop-by-hop basis, matching the semantics if
TCP were used. The basic principle is that each message, regardless
of if it carries a request or responses, will get an ACK and be
reliably retransmitted. The receiver's job is very simple, limited
to just sending ACKs. All the complexity is at the sender side. This
allows the sending implementation to trade off performance versus
implementation complexity without affecting the wire protocol.</t>
<section title="Message Format">
<t>Each message being sent is prepended with a header that
indicates the 24 bit sequence number.</t>
<figure>
<!-- struct {
uint24 seq;
} msg_data;
struct {
uint24 ack_seq;
uint32 received;
} msg_ack;
enum {data_type (1), ack_type (2)} msg_type;
select {
case data_type:
msg_data data;
case ack_type:
msg_ack ack;
} msg_thing;
public struct {
msg_thing datagram;
} datagram_thing;
-->
<artwork><![CDATA[
STRUCTURE: header
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 2
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| 0x01 | Seq |
004 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
]]></artwork>
</figure>
<t><list style="hanging">
<t hangText="Seq">The sequence number of the message.</t>
</list></t>
<t>Each DTLS session has it own sequence number. Initially the
value is zero and it increments by exactly one for each message
sent over that DTLS session.</t>
</section>
<section title="Acknowledgement Format">
<figure>
<artwork><![CDATA[
STRUCTURE: ACK
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 2
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| 0x02 | Ack Seq |
004 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Received |
008 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
]]></artwork>
</figure>
<t>When the receiver receive a message, it SHOULD immediately send
an ACK message. The receiver MUST keep track of the 32 most recent
sequence numbers received on this DTLS flow. The contents of this
packet are:</t>
<t><list style="hanging">
<t hangText="Ack Seq -">The sequence number of the message
being acknowledged.</t>
<t hangText="Received -">A bitmask indicating whether or not
each of the previous 32 packets has been received. The high
order bit represents the first packet in the sequence
space.</t>
</list></t>
<t>The received field bits in the ACK provide a very high degree
of redundancy for the sender to figure out which packets the
receiver received and can then estimate packet loss rates. If the
sender also keeps track of the time at which recent sequence
numbers were sent, the RTT can be estimated.</t>
</section>
<section title="Retransmission and Flow Control">
<t>Because the receiver's role is limited to providing packet
acknowledgements, a wide variety of congestion control algorithms
can be implemented on the sender side while using the same basic
wire protocol. It is RECOMMENDED that senders implement use
TFRC-SP<xref target="RFC4828"></xref> and use the received bitmask
to allow the sender to compute packer loss event rates. Senders
MUST implement a retransmission and congestion control scheme no
more aggressive then TFRC-SP.</t>
</section>
</section>
</section>
<section title="HIP">
<t>RELOAD MAY also be used with a HIP transport using the architecture
for HIP BONE described in <xref
target="I-D.camarillo-hip-bone"></xref>. From the perspective of the
P2P layer, HIP looks very much like normal IP. Either TLS (over TCP)
or DTLS (over UDP) is run over top of the HIP. Thus the reliability
and congestion control schemes are the same for DTLS section. If an
overlay is configured such that HIP is the only transport that it will
use, then it may make sense to configure the p2p layer to only offer
the ORCHID when gather candidate addresses for ICE. This will
effectively disable ICE at the p2p layer.</t>
<t>For overlays that use HIP, the enrollment server MUST provide each
peer with a unique ORCHID and use that ORCHID to generate the peer-id
for the peer (see <xref target="sec-credentials"></xref>. Later when
the HIP layer wishes to tunnel a message (such as an I1 message)
through the overlay, the HIP layer can use the ORCHID to generate the
peer-id, and then use the TUNNEL message with the HIP to route the
message to that the peer that owns that ORCHID.</t>
</section>
</section>
<section anchor="sec-requests" title="Method Definitions">
<t>In this section, we define the initial set of methods supported by
RELOAD. New methods are defined by adding new method codes. Each method
defines the contents of the payload element (see <xref
target="sec-payload"></xref>).</t>
<t>PDUs are named using the following convention. For method type FOO,
the request PDU is named FOO_Q and the response PDU is named FOO_A (as
are the method codes). When discussing the PDU itself, we use these
terms. Throughout the rest of the document we refer to the FOO method or
the FOO request/response for easier readability.</t>
<section title="Connection Management">
<section title="PING">
<t>PING is used to test connectivity along a path. A ping can be
addressed to a specific peer-id or to the broadcast peer-id (all
1s). In either case, the target peer-ids respond with a simple
response containing some status information.</t>
<section title="Request Definition">
<t>The PING_Q message contains a list (potentially empty) of the
pieces of status information that the requester would like the
responder to provide.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
uint8 ping_info<255>;
} ping_q;
]]></artwork>
</figure>
<!--
ping-request = [ping-request-info-responsible]
[ping-request-info-num_resources] (uint8)*
ping-request-info-responsible = 0x01
ping-request-info-num_resource = 0x02
-->
<t>The two currently defined types are:</t>
<figure>
<artwork><![CDATA[
RESPONSIBLE-SET : 0x01
NUM-RESOURCES : 0x02
]]></artwork>
</figure>
<t>RESPONSIBLE-SET indicates that the peer should Respond with the
fraction of the overlay for which the responding peer is
responsible (in parts per billion).</t>
<t>NUM-RESOURCES indicates that the peer should Respond with the
number of resources currently being stored by the peer.</t>
</section>
<section title="Response Definition">
<t>A successful PING_A response contains the information elements
requested by the peer.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
uint64 response_id;
ping_info_data infos<65000>;
} ping_a;
]]></artwork>
</figure>
<t>A PING_A message contains the following elements:</t>
<t><list style="hanging">
<t hangText="Response ID -">A randomly generated 64-bit
response ID. This is used to distinguish PING responses in
cases where the PING request is multicast.</t>
<t hangText="Infos -">A sequence of ping info data structures,
as shown below.</t>
</list></t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
enum {info_responsible_type(1),
info_num_resources_type(2)}
ping_info_types;
public struct {
ping_info_select ping_info_data;
} ping_info_data;
select {
case info_responsible_type:
uint32 responsible_ppb;
case info_num_resources_type:
uint32 num_resources;
} ping_info_select;
]]></artwork>
</figure>
<t>The ping info data elements are simple typed elements, with a
type identifier as the leading 16 bits and then arbitrary
(type-specific) text following. In the case of the two defined
types, the responses are 32-bit integers.</t>
<!--
ping-response = ping-response-id (ping-response-value)*
ping-response-value = ping-response-info-responsible /
ping-response-info-num-resources /
ping-response-info-any
ping-response-info-responsible = ping-request-info-responsible
0x04
uint32 ; parts per billion
ping-response-info-num-resources = ping-request-info-num-resources
0x04
uint32 ; count
ping-response-info-any = uint8
uint8
*byte
-->
<t>The responding peer SHOULD include any values that the
requesting peer requested and that it recognizes. They SHOULD be
returned in the requested order.</t>
</section>
</section>
<section anchor="sec-connect-details" title="CONNECT">
<t>A node sends a CONNECT request when it wishes to establish a
direct TCP or UDP connection to another node for the purposes of
sending RELOAD messages or application layer protocol messages, such
as SIP. Detailed procedures for the CONNECT and its response are
described in <xref target="sec-connect-ice"></xref>.</t>
<t><list style="hanging">
<t hangText="Note:">A CONNECT does not result in updating the
routing table of either node. That function is performed by
UPDATEs. If node A has CONNECTed to node B, it MAY route
messages which are directly addressed to B through that channel
but MUST NOT route messages through B to other peers via that
channel.</t>
</list></t>
<section anchor="sec-connect-request" title="Request Definition">
<t>A CONNECT_Q message contains the requesting peer's ICE
connection parameters formatted into a binary structure.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
string ufrag;
string password;
uint16 application;
string fingerprint;
string role;
candidate candidate_list<65000>;
} connect_data;
public struct {
string candidate_string;
} candidate;
]]></artwork>
</figure>
<!--
connect-data = connect-ufrag
connect-password
connect-application
connect-fingerprint
connect-actpass
1*connect-attribute
connect-ufrag = string ; ICE "password"
connect-password = string ; ICE "ufrag"
connect-application = uint16 ; an IANA registered port number
connect-fingerprint = string ; RFC 4572 "fingerprint"
connect-actpass = string ; RFC 4145 "role"
;
connect-attribute = string ; ICE candidate
-->
<t>The values contained in connect-request are:</t>
<t><list style="hanging">
<t hangText="Ufrag -">The username fragment (from ICE)</t>
<t hangText="Password -">The ICE password.</t>
<t hangText="Application -">A 16-bit port number. This port
number represents the IANA registered port of the protocol
that is going to be sent on this connection. For SIP, this is
5060 or 5061, and for RELOAD is TBD. By using the IANA
registered port, we avoid the need for an additional registry
and allow RELOAD to be used to set up connections for any
existing or future application protocol.</t>
<t hangText="Fingerprint -">One fingerprint attribute (from
RFC 4572 <xref target="RFC4572"></xref>).</t>
<t hangText="Role -">An active/passive/actpass attribute from
RFC 4145 <xref target="RFC4145"></xref>.</t>
<t hangText="Candidate -">One or more ICE candidate values.
Each candidate has an IP address, IP address family, port,
transport protocol, priority, foundation, component ID, STUN
type and related address. The candidate_list is a list of
string candidate values.</t>
</list></t>
<t>These values should be generated using the procedures of <xref
target="sec-connect-ice"></xref>.</t>
</section>
<section anchor="sec-connect-response" title="Response Definition">
<t>If a peer receives a CONNECT request, it SHOULD follow the
procedures of <xref target="sec-connect-ice"></xref> to process
the request and generate its own response (a CONNECT_A) containing
a connect_data object. It should then begin ICE checks. When a
peer receives a CONNECT response, it SHOULD parse the response and
begin its own ICE checks.</t>
</section>
</section>
<section anchor="sec-tunnel-details" title="TUNNEL">
<t>A node sends a TUNNEL request when it wishes to exchange
application-layer protocol messages without the expense of
establishing a direct connection via CONNECT or when ICE is unable
to establish a direct connection via CONNECT and a TURN relay is not
available. The application-level protocols that are routed via the
TUNNEL request are defined by that application's usage.</t>
<t><list style="hanging">
<t hangText="Note:">The decision of whether to route
application-level traffic across the overlay or to open a direct
connection requires careful consideration of the overhead
involved in each transaction. Establishing a direct connection
requires greater initial setup costs, but after setup,
communication is faster and imposes no overhead on the overlay.
For example, for the SIP usage, an INVITE request to establish a
voice call might be routed over the overlay, a SUBSCRIBE with
regular updates would be better used with a CONNECT, and media
would both impose too great a load on the overlay and likely
receive unacceptable performance. However, there may be a
tradeoff between locating TURN servers and relying on TUNNEL for
packet routing.</t>
</list></t>
<t>When a usage requires the TUNNEL method, it must specify the
specific application protocol(s) that will be TUNNELed and for each
protocol, specify:</t>
<t><list style="symbols">
<t>An application attribute that indicates the protocol being
tunneled. This the IANA-registered port of the application
protocol.</t>
<t>The conditions under which the application will be TUNNELed
over the overlay rather than using a direct CONNECT.</t>
<t>A mechanism for moving future application-level communication
from TUNNELing on the overlay to a direct CONNECTion, or an
explanation why this is unnecessary.</t>
<t>A means of associating messages together as required for
dialog-oriented or request/response-oriented protocols.</t>
<t>How the TUNNELed message (and associated responses) will be
delivered to the correct application. This is particularly
important if there might be multiple instances of the
application on or behind a single peer.</t>
</list></t>
<section anchor="sec-tunnel-request" title="Request Definition">
<t>The TUNNEL_Q message contains the application PDU that the
requesting peer wishes to transmit, along with some control
information identifying the handling of the PDU.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
uint16 application;
string dialog_id;
string application_pdu;
} tunnel_q;
]]></artwork>
</figure>
<t>The values contained in the TUNNEL_Q are:</t>
<t><list style="hanging">
<t hangText="Application -">A 16-bit port number. This port
number represents the IANA registered port of the protocol
that is going to be sent on this connection. For SIP, this is
5060 or 5061, and for RELOAD is TBD. By using the IANA
registered port, we avoid the need for an additional registry
and allow RELOAD to be used to set up connections for any
existing or future application protocol.</t>
<t hangText="Dialog ID -">An arbitrary string providing an
application-defined way of associating related TUNNELed
messages. This attribute may also encode sequence information
as required by the application protocol.</t>
<t hangText="Application PDU -">An application PDU in the
format specified by the application.</t>
</list></t>
</section>
<section anchor="sec-tunnel-response" title="Response Definition">
<t>A TUNNEL_A message serves as confirmation that the message was
received by the destination peer. It implies nothing about the
processing of the application. If the application protocol
specifies an acknowledgement or confirmation, that must be sent
with a separate TUNNEL request</t>
</section>
</section>
</section>
<section title="Data Storage and Retrieval">
<t>The STORE, FETCH, and REMOVE methods are used to manipulate
information in the DHT. They form an instantiation of the abstract GET
and PUT operations described in <xref
target="I-D.ietf-p2psip-concepts"></xref>.</t>
<section anchor="sec-store" title="STORE">
<t>The STORE method is used to store data in the overlay. As
described in <xref target="sec-structured-data"></xref>, each
location may contain data of multiple kinds. Each kind-id is a code
point assigned to a specific application usage by IANA. As part of
the Usage definition, protocol designers may define constraints,
such as limits on size, on the values which may be stored. For many
kinds, the set may be restricted to a single item; some sets may be
allowed to contain multiple identical items while others may only
have unique items. The protocol currently defines the following data
models:</t>
<t><list style="symbols">
<t>single value</t>
<t>array</t>
<t>dictionary</t>
</list></t>
<t>Each kind MUST specify the appropriate data model for that kind.
The format of the STORE request depends on the data model.</t>
<section anchor="sec-store-req" title="Request Definition">
<t>A STORE_Q message is a sequence of kind-data pairs, each of
which represents a sequence of stored values for a given kind. The
same kind-id MUST NOT be used twice in a given store request. Each
value is then processed in turn. These operations MUST be atomic.
If any operation fails, the state MUST be rolled back to before
the request was received.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
resource_id resource;
store_kind_data store_data<4000000000>;
} store_q;
]]></artwork>
</figure>
<t>A single STORE request stores data of a number of kinds to a
single resource location. The contents of the request are:</t>
<t><list style="hanging">
<t hangText="Resource -">The resource to store at.</t>
<t hangText="Store Kind Data -">A series of elements, one for
each kind of data to be stored.</t>
</list></t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
kind_id kind;
uint8 data_model;
generation_counter generation;
stored_data values<65000>;
} store_kind_data;
]]></artwork>
</figure>
<t>Each store kind data element represents the data to be stored
for a single kind-id. The contents of the element are:</t>
<t><list style="hanging">
<t hangText="Kind -">The kind-id. Implementations SHOULD
reject requests corresponding to unknown kinds unless
specifically configured otherwise.</t>
<t hangText="Data Model -">The data model of the data.</t>
<t hangText="Generation -">The expected current state of the
generation counter (approximately the number of times this
object has been written, see below for details).</t>
<t hangText="Values -">The value or values to be stored. This
may contain one or more stored_data values depending on the
data model associated with each kind.</t>
</list></t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
uint32 length;
uint64 storage_time;
uint32 lifetime;
blob data-value;
blob signature;
} stored_data;
]]></artwork>
</figure>
<t>Each stored_data element represents a single stored data value.
These elements are individually signed. The contents of the
element are as follows:</t>
<t><list style="hanging">
<t hangText="Length -">The length of the stored data
element.</t>
<t hangText="Storage Time -">The time when the data was stored
in absolute time, represented in seconds since the Unix epoch.
Any attempt to store a data value with a storage time before
that of a value known to the receiving peer MUST generate a
412 error. This prevents rollback attacks. Note that this does
not require synchronized clocks: the receiving peer uses the
storage time in the previous store, not its own clock.</t>
<t hangText="Lifetime -">The validity period for the data, in
seconds, starting from the time of store.</t>
<t hangText="Signature -">A signature over the data value.
<xref target="sec-data-sig"></xref> describes the signature
computation. The element is formatted as described in <xref
target="sec-signature"></xref></t>
<t hangText="Data Value -">The data value itself, as described
below.</t>
</list></t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
opaque value<65000>;
} single_value_entry;
public struct {
int32 index;
opaque value<65000>;
} array_entry;
public struct {
opaque key<65000>;
opaque value<65000>;
} dictionary_entry;
]]></artwork>
</figure>
<!-- store-request = resource-id 1*store-type-data
store-type-data = type-id
generation-counter
uint16 ; number of stored data elements
1*stored-data
generation-counter = uint64
stored-data = uint32 ; length of the rest of the production
stored-data-expiry-time
stored-data-value
signature
stored-data-value = single-value / array-entry / dictionary-entry /
*byte
single-value = data-value
data-value = uint32 ; length
*byte ; data
array-entry = int32 ; index
data-value
dictionary-entry = data-value ; key
data-value ; value
-->
<t>The responsible peer MUST perform the following checks:</t>
<t><list style="symbols">
<t>The kind-id is known.</t>
<t>The signature over the message is valid or (depending on
overlay policy) no signature is required.</t>
<t>The signatures over each individual data element (if any)
are valid.</t>
<t>Each element is signed by a credential which is authorized
to write this kind at this resource-id</t>
<t>If the generation-counter is non-zero, it must equal the
current value of the generation-counter for this kind. This
feature allows the generation counter to be used in a way
similar to the HTTP Etag feature.</t>
<t>The storage time values are greater than that of any value
which would be replaced by this STORE. [[OPEN ISSUE: do peers
need to save the storage time of REMOVEs to prevent
reinsertion?]]</t>
</list></t>
<t>If all these checks succeed, the peer MUST attempt to store the
data values. If the store succeeds and the data is changed, then
the peer must increase the generation counter by at least one. If
there are multiple stored values in a single store_kind_data, it
is permissible for the peer to increase the generation counter by
only 1 for the entire kind-id, or by 1 or more than one for each
value.</t>
<t>We now discuss each data model:</t>
<section title="Single Value">
<t>There may be only one single-value element for each
resource-id, kind-id pair. A store of a new single-value element
MUST overwrite the current value.</t>
</section>
<section title="Array">
<t>A store of an array entry replaces (or inserts) the given
value at the location specified by the index. Arrays are
zero-based. Note that arrays can be sparse. Thus, a store of "X"
at index 2 in an empty array produces an array with the values [
NA, NA, "X"]. Future attempts to fetch elements at index 0 or 1
will return empty strings. If the index value is -1, then the
value is placed at the end of the array.</t>
</section>
<section title="Dictionary">
<t>A stored dictionary entry has a dictionary-key used as a
lookup key and a dictionary-value containing the data. There may
be only one value for any given dictionary-key and therefore a
write to a dictionary-key overwrites whatever is there.</t>
</section>
</section>
<section title="Response Definition">
<t>In response to a successful STORE request the peer MUST return
a STORE_A message containing a series of store_kind_response
elements containing the current value of the generation counter
for each kind-id, as well as a list of the peers where the data
was replicated.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
kind_id kind;
generation_counter generation;
peer_id replicas<65535>;
} store_kind_response;
]]></artwork>
</figure>
<!--
store-response = 1*store-type-response
store-type-response = type-id generation-counter
-->
<t>The contents of each element are:</t>
<t><list style="hanging">
<t hangText="Kind -">The kind-id being represented.</t>
<t hangText="Generation -">The current value of the generation
counter for that kind-id.</t>
<t hangText="Replicas -">The list of other peers at which the
data was/will-be replicated. In DHTs and applications where
the responsible peer is intended to store redundant copies,
this allows the storing peer to independently verify that the
replicas were in fact stored.</t>
</list></t>
<t>The response itself is just the store_kind_response values
packed end-to-end.</t>
<t>If the request was rejected because of an invalid generation
counter, then the store-response MUST also be returned, but with a
response code of 412. Otherwise, the response MAY contain a
response-error-reason production or MAY be empty. [[TODO: The
generation counter may need more thinking for uniqueness.]]</t>
</section>
<section anchor="sec-data-sig" title="Data Signature Computation">
<t>Each stored-data element is individually signed. However, the
signature also must be self-contained and cover the kind-id and
resource-id even though they are not present in the stored value.
The data signed is defined as:</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
resource_id resource;
kind_id kind;
blob stored_data;
} stored_data_to_be_signed;
]]></artwork>
</figure>
<!-- stored-data-to-be-signed = resource-id type-id stored-data -->
<t>The contents of this value are as follows:</t>
<t><list style="hanging">
<t hangText="Resource -">The resource ID where this data is
stored.</t>
<t hangText="Type -">The kind-id for this data.</t>
<t hangText="Stored Data -">The contents of the stored data
value, as described in the stored_data PDU of <xref
target="sec-store-req"></xref></t>
</list></t>
<t>[[TODO: Should we include the identity?.]]</t>
<t>Once the signature has been computed, the signature is
represented using a signature element, as described in <xref
target="sec-signature"></xref>.</t>
</section>
</section>
<section title="FETCH">
<t>The FETCH request retrieves one or more data elements stored at a
given resource-id.</t>
<section title="Request Definition">
<t>A FETCH_Q message consists of a single fetch_request element
followed by a series of fetch_kind_data elements.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
resource_id resource;
fetch_kind_data fetch_data<65000>;
} fetch_q;
]]></artwork>
</figure>
<t>The contents of the request are as follows:</t>
<t><list style="hanging">
<t hangText="Resource -">The resource ID to fetch from.</t>
<t hangText="Fetch Data -">A sequence of data specifiers, one
for each desired kind-id.</t>
</list></t>
<t>Each fetch_kind_data element is specified as follows.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
kind_id kind;
uint8 data_model;
generation_counter generation;
fetch_data_reference reference;
} fetch_kind_data;
]]></artwork>
</figure>
<t><list style="hanging">
<t hangText="Kind -">The kind-id of the data being fetched.
Implementations SHOULD reject requests corresponding to
unknown kinds unless specifically configured otherwise.</t>
<t hangText="Data Model -">The data model of the data.</t>
<t hangText="Generation -">The last generation counter that
the requesting peer saw. This is used to avoid unnecessary
fetches.</t>
<t hangText="Reference -">A reference to the data value being
requested within the data model specified for the kind, as
specified below.</t>
</list></t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
int32 first;
int32 last;
} fetch_array_reference;
public struct {
dictionary_key dictionary_keys<65000>;
} fetch_dictionary_reference;
public struct {
string key_value<65000>;
} dictionary_key;
]]></artwork>
</figure>
<!--
enum {fetch_single (1), fetch_array (2), fetch_dictionary (3)} fetch_model;
select {
case fetch_array:
fetch_array_reference array_ref;
case fetch_dictionary:
fetch_dictionary_reference dictionary_ref;
} fetch_data_reference;
-->
<!-- fetch-request = resource-id 1*fetch-type-data
fetch-type-data = type-id
generation-counter
fetch-data-reference
fetch-data-reference = array-reference / dictionary-reference / *byte
array-reference = int32 int32
dictionary-reference = uint32 ; number of keys
*data-value ; list of dictionary keys-->
<t>As with STORE, the FETCH_Q contains a list of kind-ids and
associated references. The reference encoding depends on the kind
of value being stored.</t>
<t><list style="symbols">
<t>If the data is of data model single value, the reference is
empty.</t>
<t>If the data is of data model array, the reference contains
two integers. The first integer is the beginning of the range
and the second is the end of the range. 0 is used to indicate
the first element and -1 is used to indicate the final
element. The beginning of the range MUST be earlier in the
array then the end.</t>
<t>If the data is of data model dictionary then the reference
contains a list of the dictionary keys being requested. If no
keys are specified, than this is a wildcard fetch and all
key-value pairs are returned. [[TODO: We really need a way to
return only the keys. We'll need to modify this.]]</t>
</list></t>
<t>The generation-counter is used to indicate the requester's
expected state of the storing peer. If the generation-counter in
the request matches the stored counter, then the storing peer
returns a cache hit indicator rather than the stored data.</t>
<t>Note that because the certificate for a user is typically
stored at the same location as any data stored for that user, a
requesting peer which does not already have the user's certificate
should request the certificate in the FETCH as an
optimization.</t>
</section>
<section title="Response Definition">
<t>The response to a successful FETCH request is a FETCH_A message
containing the data requested by the requester.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
kind_id kind;
generation_counter generation;
opaque stored_data<65000>;
} fetch_a;
]]></artwork>
</figure>
<!--
fetch-response = 1*fetch-type-value
fetch-type-value = type-id
generation-counter
uint16 ; number of stored data elements
1*stored-data
-->
<t>There MUST be one fetch_kind_data element for each kind-id in
the request. If the generation-counter in the request matches the
generation-counter in the stored data, then the count of stored
data elements MUST be zero. Otherwise, all relevant data values
MUST be returned. A nonexistent value is represented as a value
with an empty data value portion and no signature. In particular,
if a dictionary key that does not exist is requested, then there
must be a dictionary entry with that key but an empty value.</t>
</section>
</section>
<section title="REMOVE">
<t>The REMOVE request is used to remove a stored element or elements
from the storing peer. Although each kind defines its own access
control requirements, in general only the original signer of the
data should be allowed to remove it. Any successful remove of an
existing element for a given kind MUST increment the generation
counter by at least one.</t>
<t>A remove-request has exactly the same syntax as a FETCH request
except that each entry represents a set of values to be removed
rather than returned. The same kind-id MUST NOT be used twice in a
given remove-request. Each fetch_kind_data is then processed in
turn. These operations MUST be atomic. If any operation fails, the
state MUST be rolled back to before the request was received.</t>
<t>Before processing the REMOVE request, the peer MUST perform the
following checks.</t>
<t><list style="symbols">
<t>The kind-id is known.</t>
<t>The signature over the message is valid or (depending on
overlay policy) no signature is required.</t>
<t>The signer of the message has permissions which permit him to
remove this kind of data.</t>
<t>If the generation-counter is non-zero, it must equal the
current value of the generation-counter for this kind. This
feature allows the generation counter to be used in a way
similar to the HTTP Etag feature.</t>
</list></t>
<t>Assuming that the request is permitted, the operations proceed as
follows.</t>
<section title="Single Value">
<t>A REMOVE of a single value element simple causes it not to
exist. If no such element exists, then this simply is a silent
success.</t>
</section>
<section title="Array">
<t>A REMOVE of an array element (or element range) replaces those
elements with empty elements. Note that this does not cause the
array to be packed. An array which contains ["A", "B", "C"] and
then has element 0 removed produces an array containing [NA, "B",
"C"]. Note, however, that the removal of the final element of the
array shortens the array, so in the above case, the removal of
element 2 makes the array ["A", "B"].</t>
</section>
<section title="Dictionary">
<t>A REMOVE of a dictionary element (or elements) replaces those
elements with empty elements. If no such elements exist, then this
is a silent success.</t>
</section>
<section title="Response Definition">
<t>The response to a successful REMOVE simply contains a list of
the new generation counters for each kind-id, using the same
syntax as the response to a STORE request. Note that if the
generation counter does not change, that means that the requested
items did not exist. However, if the generation counter does
change, that does not mean that the items existed.</t>
</section>
</section>
<section title="FIND">
<t>The FIND request is used to explore the DHT. A FIND request for a
resource-id R and a kind-id T retrieves the resource-id (if any) of
the resource of kind T known to the target peer which is closes to
R. This method can be used to walk the DHT by interactively fetching
R_n+1=nearest(1 + R_n).</t>
<section title="Request Definition">
<t>The FIND_Q message contains a series of resource-IDs and
kind-ids identifying the resource the peer is interested in.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
resource_id resource;
kind_id ids<255>;
} find_q;
]]></artwork>
</figure>
<!-- find-request = resource-id 1*type-id -->
</section>
<t>The request contains a list of kind-ids which the FIND is for, as
indicated below.</t>
<t><list style="hanging">
<t hangText="Resource -">The desired resource-id</t>
<t hangText="Ids -">The desired kind-ids. Each value MUST only
appear once.</t>
</list></t>
<section title="Response Definition">
<t>A response to a successful FIND request is a FIND_A message
containing the closest resource-ID for each kind specified in the
request.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
kind_id kind;
resource_id closest;
} find_kind_data;
]]></artwork>
</figure>
<t>If the processing peer is not responsible for the specified
resource-id, it SHOULD return a 404 error.</t>
<t>When each kind is defined, it can indicate if the kind is not
allowed to be used in a FIND request. This would be done to help
achieve some types of security properties for the data stored in
that kind.</t>
<t>For each kind-id in the request the response MUST contain a
find_response_value indicating the closest resource-id for that
kind-id unless the kind is not allowed to be used with FIND in
which case a find_kind_data for that kind_id MUST NOT be included
in the response. If a kind-id is not known, then the corresponding
resource-id MUST be 0. Note that different kind-ids may have
different closest resource-ids.</t>
<t>The response is simply a series of find_kind_data elements, one
per kind, concatenated end-to-end. The contents of each element
are:</t>
<t><list style="hanging">
<t hangText="Kind -">The kind-id.</t>
<t hangText="Closest -">The closest resource ID to the
specified resource ID. This is 0 if no resource ID is
known.</t>
</list></t>
<t>Note that the response does not contain the contents of the
data stored at these resource-ids. If the requester wants this, it
must retrieve it using FETCH.</t>
</section>
</section>
</section>
<section anchor="sec-dht-maint" title="DHT Maintenance">
<t>This section describes methods that are expected to be useful for
all DHTs. These methods have generic semantics (join, leave, update)
and some common fields, but where appropriate allow room for
DHT-specific data.</t>
<section title="JOIN">
<t>A new peer (but which already has credentials) uses the JOIN_Q
message to join the DHT. The JOIN_Q is sent to the peer which
previously was responsible for the resource-id corresponding to the
peer-id which the new peer has. This notifies the responsible peer
that the new peer is taking over some of the overlay and it needs to
synchronize its state.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
peer_id desired_peer_id;
blob dht_specific_data;
} join_q;
]]></artwork>
</figure>
<!-- join-request = peer-id -->
<t>The default JOIN_Q contains only the peer-id which the sending
peer wishes to assume. DHTs MAY specific other data to appear in
this request.</t>
<t>The responding peer responds with success or failure. However, if
it is success it MUST follow up by executing the right sequence of
STOREs and UPDATEs to transfer the appropriate section of the
overlay space to the joining peer. In addition, DHTs MAY define data
to appear in the response payload.</t>
</section>
<section title="LEAVE">
<t>The LEAVE_Q message is used to indicate that a peer is exiting
the overlay. The peer SHOULD send this message to each peer with
which it is directly connected prior to exiting the overlay.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
peer_id leaving_peer_id;
blob dht_specific_data;
} leave_q;
]]></artwork>
</figure>
<!-- leave-request = peer-id -->
<t>The default LEAVE_Q contains only the peer-id of the leaving
peer. DHTs MAY specific other data to appear in this request.</t>
<t>Upon receiving a LEAVE request, a peer MUST update its own
routing and routing table, and send the appropriate STORE/UPDATE
sequences to re-stabilize the overlay.</t>
</section>
<section title="UPDATE">
<t>Update is the primary DHT-specific maintenance message. It is
used by the sender to notify the recipient of the sender's view of
the current state of the overlay and it is up to the recipient to
take whatever actions are appropriate to deal with the state
change.</t>
<t>The contents of the UPDATE_Q message are completely DHT-specific.
The UPDATE_A response is expected to be either success or an
error.</t>
</section>
<section anchor="sec-route-query" title="ROUTE_QUERY">
<t>The ROUTE_QUERY request allows the sender to ask a peer where
they would route a message directed to a given destination. In other
words, a ROUTE-QUERY for destination X requests the peer-id where
the receiving peer would next route to get to X. A ROUTE-QUERY can
also request that the receiving peer initiate an UPDATE request to
transfer his routing table.</t>
<t>One important use of the ROUTE-QUERY request is to support
iterative routing. The way that his works is that the sender selects
one of the peers in its neighbor table and sends it a ROUTE-QUERY
message with the destination_object set to the peer-id/resource-id
it wishes to route to. The neighbor responds with the next peer-id
to send to. The sending peer then CONNECTs to that peer and repeats
the ROUTE-QUERY. Eventually, the sender gets a response from a peer
containing a peer-id that is the same as that peer. At that point,
the sender can send whatever request is needed directly to that
peer.</t>
<t>Note that this procedure only works well if all the peers are
mutually directly reachable--either by all having public IP
addresses or at least by all being behind the same NAT. Accordingly,
peers MUST only use this method if permitted by the overlay
configuration (see <xref target="sec-configuration"></xref>).</t>
<section title="Request Definition">
<t>A ROUTE_QUERY_Q message indicates the peer or resource that the
requesting peer is interested in. It also contains a "send_update"
option allowing the requesting peer to request a full copy of the
other peer's routing table.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
char send_update;
blob destination_object;
} route_query_q;
]]></artwork>
</figure>
<t>The contents of the ROUTE_QUERY_Q message are as follows:</t>
<t><list style="hanging">
<t hangText="send_update:">A single byte. This may be set to 1
to indicate that the requester wishes the responder to
initiate an UPDATE request immediately. Otherwise, this value
MUST be set to zero.</t>
<t hangText="destination_object:">The destination which the
requester is interested in. This may be any valid destination
object, including a peer-id, compressed ids, or
resource-id</t>
</list></t>
</section>
<section title="Response Definition">
<t>A response to a successful ROUTE_QUERY request is a
ROUTE_QUERY_A message containing the address of the peer to which
the responding peer would have routed the request message in
recursive routing.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
peer_id next_peer;
blob next_addr;
} route_query_a;
]]></artwork>
</figure>
<t>The contents of the ROUTE_QUERY_A are as follows:</t>
<t><list style="hanging">
<t hangText="next_peer:">The peer to which the responding peer
would route the message to in order to deliver it to the
destination listed in the request.</t>
<t hangText="next_addr:">The address of the next peer.</t>
</list></t>
<t>If the requester set the send_update flag, the responder SHOULD
initiate an UPDATE immediately after.</t>
</section>
</section>
</section>
</section>
<section anchor="sec-connect-ice" title="ICE and Connection Formation">
<t>At numerous times during the operation of RELOAD, a node will need to
establish a connection to another node. This may be for the purposes of
building finger tables when the node joins the P2P network, or when the
node learns of a new neighbor through an UPDATE and needs to establish a
connection to that neighbor.</t>
<t>In addition, a node may need to connect to another node for the
purposes of an application connection. In the case of SIP, when a node
has looked up the target AOR in the DHT, it will obtain a Node-ID that
identifies that peer. The next step will be to establish a "direct"
connection for the purposes of performing SIP signaling.</t>
<t>In both of these cases, the node starts with a destination Node-ID,
and its objective is to create a connection (ideally using TCP, but
falling back to UDP when it is not available) to the node with that
given Node-ID. The establishment of this connection is done using the
CONNECT request in conjunction with ICE. It is assumed that the reader
has familiarity with ICE.</t>
<t>RELOAD implementations MUST implement full ICE. Because RELOAD always
tries to use TCP and then UDP as a fallback, there will be multiple
candidates of the same IP version, which requires full ICE.</t>
<section title="Overview">
<t>To utilize ICE, the CONNECT method provides a basic offer/answer
operation that exchanges a set of candidates for a single "stream". In
this case, the "stream" refers not to RTP or other types of media, but
rather to a connection for RELOAD itself or for SIP signaling. The
CONNECT request contains the candidates for this stream, and the
CONNECT response contains the corresponding answer with candidates for
that stream. Though CONNECT provides an offer/answer exchange, it does
not actually carry or utilize Session Description Protocol (SDP)
messages. Rather, it carries the raw ICE parameters required for ICE
operation, and the ICE spec is utilized as if these parameters had
actually been used in an SDP offer or answer. In essence, ICE is
utilized by mapping the CONNECT parameters into an SDP for the
purposes of following the details of ICE itself. That avoids the need
for RELOAD to respecify ICE, yet allows it to operate without the
baggage that SDP would bring.</t>
<t>In addition, RELOAD only allows for a single offer/answer exchange.
Unlike the usage of ICE within SIP, there is never a need to send a
subsequent offer to update the default candidates to match the ones
selected by ICE.</t>
<t>RELOAD and SIP always run over TLS for TCP connections and DTLS
<xref target="RFC4347"></xref> for UDP "connections". Consequently,
once ICE processing has completed, both agents will begin TLS and DTLS
procedures to establish a secure link. Its important to note that, had
a TURN server been utilized for the TCP or UDP stream, the TURN server
will transparently relay the TLS messaging and the encrypted TLS
content, and thus will not have access to the contents of the
connection once it is established. Any attack by the TURN server to
insert itself as a man-in-the-middle are thwarted by the usage of the
fingerprint mechanism of RFC 4572 <xref target="RFC4572"></xref>,
which will reveal that the TLS and DTLS certificates do not match the
ones used to sign the RELOAD messages.</t>
<t>An agent follows the ICE specification as described in <xref
target="I-D.ietf-mmusic-ice"></xref> and <xref
target="I-D.ietf-mmusic-ice-tcp"></xref> with the changes and
additional procedures described in the subsections below.</t>
</section>
<section anchor="sec-collect" title="Collecting STUN Servers">
<t>ICE relies on the node having one or more STUN servers to use. In
conventional ICE, it is assumed that nodes are configured with one or
more STUN servers through some out-of-band mechanism. This is still
possible in RELOAD but RELOAD also learns STUN servers as it connects
to other peers. Because all RELOAD peers implement ICE and use STUN
keepalives, every peer is a STUN server<xref
target="I-D.ietf-behave-rfc3489bis"></xref>. Accordingly, any peer you
know about will be willing to be a STUN server for you -- though of
course it may be behind a NAT.</t>
<t>A peer on a well-provisioned wide-area overlay will be configured
with one or more bootstrap peers. These peers make an initial list of
STUN servers. However, as the peer forms connections with additional
peers, it builds more peers it can use as STUN servers.</t>
<t>Because complicated NAT topologies are possible, a peer may need
more than one STUN server. Specifically, a peer that is behind a
single NAT will typically observe only two IP addresses in its STUN
checks: its local address and its server reflexive address from a STUN
server outside its NAT. However, if there are more NATs involved, it
may discover that it learns additional server reflexive addresses
(which vary based on where in the topology the STUN server is). To
maximize the chance of achieving a direct connection, A peer SHOULD
group other peers by the peer-reflexive addresses it discovers through
them. It SHOULD then select one peer from each group to use as a STUN
server for future connections.</t>
<t>Only peers to which the peer currently has connections may be used.
If the connection to that host is lost, it MUST be removed from the
list of stun servers and a new server from the same group SHOULD be
selected.</t>
<t>OPEN ISSUE: should the peer try to keep at least one peer in each
group, even if it has no other reason for the connection? Need to
specify when to stop adding new groups if the peer is behind a really
bad NAT.</t>
<t>OPEN ISSUE: RELOAD-01 had a Peer-Info structure that allowed peers
to exchange information such as a "default" IP-port pair in UPDATEs.
This structure could be expanded to include the candidate list for a
peer, thus allowing ICE negotiation to begin or even direct
communication before a CONNECT request has been received. (The
candidate pairs for the P2P port are fixed because the same source
port is used for all connections.) However, because this would require
significant changes to the ICE algorithm, we have not introduced such
an extension at this point.</t>
</section>
<section anchor="sec-gather" title="Gathering Candidates">
<t>When a node wishes to establish a connection for the purposes of
RELOAD signaling or SIP signaling (or any other application protocol
for that matter), it follows the process of gathering candidates as
described in Section 4 of ICE <xref
target="I-D.ietf-mmusic-ice"></xref>. RELOAD utilizes a single
component, as does SIP. Consequently, gathering for these "streams"
requires a single component.</t>
<t>An agent MUST implement ICE-tcp <xref
target="I-D.ietf-mmusic-ice"></xref>, and MUST gather at least one UDP
and one TCP host candidate for RELOAD and for SIP.</t>
<t>The ICE specification assumes that an ICE agent is configured with,
or somehow knows of, TURN and STUN servers. RELOAD provides a way for
an agent to learn these by querying the ring, as described in <xref
target="sec-collect"></xref> and <xref
target="sec-stun-turn-usage"></xref>.</t>
<t>The agent SHOULD prioritize its TCP-based candidates over its
UDP-based candidates in the prioritization described in Section 4.1.2
of ICE <xref target="I-D.ietf-mmusic-ice"></xref>.</t>
<t>The default candidate selection described in Section 4.1.3 of ICE
is ignored; defaults are not signaled or utilized by RELOAD.</t>
</section>
<section title="Encoding the CONNECT Message">
<t>Section 4.3 of ICE describes procedures for encoding the SDP.
Instead of actually encoding an SDP, the candidate information (IP
address and port and transport protocol, priority, foundation,
component ID, type and related address) is carried within the
attributes of the CONNECT request or its response. Similarly, the
username fragment and password are carried in the CONNECT message or
its response. <xref target="sec-connect-details"></xref> describes the
detailed attribute encoding for CONNECT. The CONNECT request and its
response do not contain any default candidates or the ice-lite
attribute, as these features of ICE are not used by RELOAD. The
CONNECT request and its response also contain a Next-Protocol
attribute, with a value of SIP or RELOAD, which indicates what
protocol is to be run over the connection. The RELOAD CONNECT request
MUST only be utilized to set up connections for application protocols
that can be multiplexed with STUN and RELOAD itself.</t>
<t>Since the CONNECT request contains the candidate information and
short term credentials, it is considered as an offer for a single
media stream that happens to be encoded in a format different than
SDP, but is otherwise considered a valid offer for the purposes of
following the ICE specification. Similarly, the CONNECT response is
considered a valid answer for the purposes of following the ICE
specification.</t>
<t>Since all messages with RELOAD are secured between nodes, the node
MUST implement the fingerprint attribute of RFC 4572 <xref
target="RFC4572"></xref>, and encode it into the CONNECT request and
response as described in <xref target="sec-connect-details"></xref>.
This fingerprint will be matched with the certificates utilized to
authenticate the RELOAD CONNECT request and its response.</t>
<t>Similarly, the node MUST implement the active, passive, and actpass
attributes from RFC 4145 <xref target="RFC4145"></xref>. However, here
they refer strictly to the role of active or passive for the purposes
of TLS handshaking. The TCP connection directions are signaled as part
of the ICE candidate attribute.</t>
</section>
<section title="Verifying ICE Support">
<t>An agent MUST skip the verification procedures in Section 5.1 and
6.1 of ICE. Since RELOAD requires full ICE from all agents, this check
is not required.</t>
</section>
<section title="Role Determination">
<t>The roles of controlling and controlled as described in Section 5.2
of ICE are still utilized with RELOAD. However, the offerer (the
entity sending the CONNECT request) will always be controlling, and
the answerer (the entity sending the CONNECT response) will always be
controlled. The connectivity checks MUST still contain the
ICE-CONTROLLED and ICE-CONTROLLING attributes, however, even though
the role reversal capability for which they are defined will never be
needed with RELOAD. This is to allow for a common codebase between ICE
for RELOAD and ICE for SDP.</t>
</section>
<section title="Connectivity Checks">
<t>The processes of forming check lists in Section 5.7 of ICE,
scheduling checks in Section 5.8, and checking connectivity checks in
Section 7 are used with RELOAD without change.</t>
</section>
<section title="Concluding ICE">
<t>The controlling agent MUST utilize regular nomination. This is to
ensure consistent state on the final selected pairs without the need
for an updated offer, as RELOAD does not generate additional
offer/answer exchanges.</t>
<t>The procedures in Section 8 of ICE are followed to conclude ICE,
with the following exceptions:</t>
<t><list style="symbols">
<t>The controlling agent MUST NOT attempt to send an updated offer
once the state of its single media stream reaches Completed.</t>
<t>Once the state of ICE reaches Completed, the agent can
immediately free all unused candidates. This is because RELOAD
does not have the concept of forking, and thus the three second
delay in Section 8.3 of ICE does not apply.</t>
</list></t>
</section>
<section title="Subsequent Offers and Answers">
<t>An agent MUST NOT send a subsequent offer or answer. Thus, the
procedures in Section 9 of ICE MUST be ignored.</t>
</section>
<section title="Media Keepalives">
<t>STUN MUST be utilized for the keepalives described in Section 10 of
ICE.</t>
</section>
<section title="Sending Media">
<t>The procedures of Section 11 apply to RELOAD as well. However, in
this case, the "media" takes the form of application layer protocols
(RELOAD or SIP for example) over TLS or DTLS. Consequently, once ICE
processing completes, the agent will begin TLS or DTLS procedures to
establish a secure connection. The fingerprint from the CONNECT
request and its response are used as described in RFC 4572 <xref
target="RFC4572"></xref>, to ensure that another node in the P2P
network, acting as a TURN server, has not inserted itself as a
man-in-the-middle. Once the TLS or DTLS signaling is complete, the
application protocol is free to use the connection.</t>
<t>The concept of a previous selected pair for a component does not
apply to RELOAD, since ICE restarts are not possible with RELOAD.</t>
</section>
<section title="Receiving Media">
<t>An agent MUST be prepared to receive packets for the application
protocol (TLS or DTLS carrying RELOAD, SIP or anything else) at any
time. The jitter and RTP considerations in Section 11 of ICE do not
apply to RELOAD or SIP.</t>
</section>
</section>
<section title="DHT Algorithms">
<section title="Generic Algorithm Requirements">
<t>When specifying a new DHT, at least the following need to be
described:</t>
<t><list style="symbols">
<t>Joining procedures, including the contents of the JOIN
message.</t>
<t>Stabilization procedures, including the contents of the UPDATE
message, the frequency of topology probes and keepalives, and the
mechanism used to detect when peers have disconnected.</t>
<t>Exit procedures, including the contents of the LEAVE
message.</t>
<t>The hash algorithm used to go from a Unhashed-ID, such as a
user name, to a Resource-ID. This also includes the length of the
Resource-IDs and Peer-IDs</t>
<t>The procedures that peers use to route messages.</t>
<t>The replication strategy used to ensure data redundancy.</t>
</list></t>
</section>
<section anchor="sec.chord-algorithm" title="Chord Algorithm ">
<t></t>
<t>This algorithm is assigned the name chord-128-2-16+ to indicate it
is based on Chord, uses a 128 bit hash function, stores 2 redundant
copies of all data, and has finger tables with at least 16
entries.</t>
<section title="Overview">
<t>The algorithm described here is a modified version of the Chord
algorithm. Each peer keeps track of a finger table of 16 entries and
a neighborhood table of 6 entries. The neighborhood table contains
the 3 peers before this peer and the 3 peers after it in the DHT
ring. The first entry in the finger table contains the peer half-way
around the ring from this peer; the second entry contains the peer
that is 1/4 of the way around; the third entry contains the peer
that is 1/8th of the way around, and so on. Fundamentally, the chord
data structure can be thought of a doubly-linked list formed by
knowing the successors and predecessor peers in the neighborhood
table, sorted by the peer-id. As long as the successor peers are
correct, the DHT will return the correct result. The pointers to the
prior peers are kept to enable inserting of new peers into the list
structure. Keeping multiple predecessor and successor pointers makes
it possible to maintain the integrity of the data structure even
when consecutive peers simultaneously fail. The finger table forms a
skip list, so that entries in the linked list can rapidly be found -
it needs to be there so that peers can be found in O(log(N)) time
instead of the typical O(N) time that a linked list would
provide.</t>
<t>A peer, n, is responsible for a particular Resource-ID k if k is
less than or equal to n and k is greater than p, where p is the peer
id of the previous peer in the neighborhood table. Care must be
taken when computing to note that all math is modulo 2^128.</t>
</section>
<section title="Routing">
<t>If a peer is not responsible for a Resource-ID k, then it routes
a request to that location by routing it to the peer in either the
neighborhood or finger table that has the largest peer-id that is in
the interval between the peer and k.</t>
</section>
<section title="Redundancy ">
<t>When a peer receives a STORE request for Resource-ID k, and it is
responsible for Resource-ID k, it stores the data and returns a
SUCCESS response. [[Open Issue: should it delay sending this SUCCESS
until it has successfully stored the redundant copies?]]. It then
sends a STORE request to its successor in the neighborhood table and
to that peers successor. Note that these STORE requests are
addressed to those specific peers, even though the Resource-ID they
are being asked to store is outside the range that they are
responsible for. The peers receiving these check they came from an
appropriate predecessor in their neighborhood table and that they
are in a range that this predecessor is responsible for, and then
they store the data.</t>
<t>Note that a malicious node can return a success response but not
store the data locally or in the replica set. Requesting peers which
wish to ensure that the replication actually occurred SHOULD contact
each peer listed in the replicas field of the STORE response and
retrieve a copy of the data. [[TODO: Do we want to have some
optimization in FETCH where they can retrieve just a digest instead
of the data values?]]</t>
</section>
<section title="Joining">
<t>The join process for a joining party (JP) with peer-id n is as
follows.</t>
<t><list style="numbers">
<t>JP connects to its chosen bootstrap node.</t>
<t>JP uses a series of PINGs to populate its routing table.</t>
<t>JP sends CONNECT requests to initiate connections to each of
the peers in the connection table as well as to the desired
finger table entries. Note that this does not populate their
routing tables, but only their connection tables, so JP will not
get messages that it is expected to route to other nodes.</t>
<t>JP enters all the peers it contacted into its routing
table.</t>
<t>JP sends a JOIN to its immediate successor, the admitting
peer (AP) for peer-id n. The AP sends the response to the
JOIN.</t>
<t>AP does a series of STORE requests to JP to store the data
that JP will be responsible for.</t>
<t>AP sends JP an UPDATE explicitly labeling JP as its
predecessor. At this point, JP is part of the ring and
responsible for a section of the overlay. AP can now forget any
data which is assigned to JP and not AP.</t>
<t>AP sends an UPDATE to all of its neighbors with the new
values of its neighbor set (including JP).</t>
<t>JP sends UPDATES to all the peers in its routing table.</t>
</list></t>
<t>In order to populate its routing table, JP sends a PING via the
bootstrap node directed at resource-id n+1 (directly after its own
resource-id). This allows it to discover its own successor. Call
that node p0. It then sends a ping to p0+1 to discover its successor
(p1). This process can be repeated to discover as many successors as
desired. The values for the two peers before p will be found at a
later stage when n receives an UPDATE.</t>
<t>In order to set up its neighbor table entry for peer i, JP simply
sends a CONNECT to peer (n+2^(numBitsInPeerId-i). This will be
routed to a peer in approximately the right location around the
ring.</t>
</section>
<section title="Routing CONNECTs">
<t>When a peer needs to CONNECT with a new peer in its neighborhood
table, it MUST source-route the CONNECT request through the peer
from which it learned the new peer's peer-id. Source-routing these
requests allows the overlay to recover from instability.</t>
<t>All other CONNECT requests, such as those for new finger table
entries, are routed conventionally through the overlay.</t>
<t>If a peer is unable to successfully CONNECT with a peer that
should be in its neighborhood, it MUST locate either a TURN server
or another peer in the overlay, but not in its neighborhood, through
which it can exchange messages with its neighbor peer</t>
</section>
<section title="UPDATEs">
<t>An UPDATE is defined as</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
peer_id predecessors<255>;
peer_id successors<255>;
} chord_update;
]]></artwork>
</figure>
<!--
chord-update = uint8 ; number of predecessors
*peer-id ; predecessor set
uint8 ; number of successors
*peer-id ; successor set
-->
<t>The contents of this message are:</t>
<t><list style="hanging">
<t hangText="Predecessors -">The predecessor set of the
UPDATEing peer.</t>
<t hangText="Successors -">The successor set of the UPDATEing
peer.</t>
</list></t>
<t>A peer MUST maintain an association (via CONNECT) to every member
of its neighbor set. A peer MUST attempt to maintain at least three
predecessors and three successors. However, it MUST send its entire
set in any UPDATE message.</t>
<section title="Sending UPDATEs">
<t>Every time a connection to a peer in the neighborhood set is
lost (as determined by connectivity pings or failure of some
request), the peer should remove the entry from its neighborhood
table and replace it with the best match it has from the other
peers in its routing table. It then sends an UPDATE to all its
remaining neighbors. The update will contain all the peer-ids of
the current entries of the table (after the failed one has been
removed). Note that when replacing a successor the peer SHOULD
delay the creation of new replicas for 30 seconds after removing
the failed entry from its neighborhood table in order to allow a
triggered update to inform it of a better match for its
neighborhood table.</t>
<t>If connectivity is lost to all three of the peers that succeed
this peer in the ring, then this peer should behave as if it is
joining the network and use PINGs to find a peer and send it a
JOIN. If connectivity is lost to all the peers in the finger
table, this peer should assume that it has been disconnected from
the rest of the network, and it should periodically try to join
the DHT.</t>
</section>
<section title="Receiving UPDATEs">
<t>When a peer, N, receives an UPDATE request, it examines the
peer-ids in the UPDATE_Q and at its neighborhood table and decides
if this UPDATE_Q would change its neighborhood table. This is done
by taking the set of peers currently in the neighborhood table and
comparing them to the peers in the update request. There are three
major cases:</t>
<t><list style="symbols">
<t>The UPDATE_Q contains peers that would not change the
neighbor set because they match the neighborhood table.</t>
<t>The UPDATE_Q contains peers closer to N than those in its
neighborhood table.</t>
<t>The UPDATE_Q defines peers that indicate a neighborhood
table further away from N than some of its neighborhood table.
Note that merely receiving peers further away does not
demonstrate this, since the update could be from a node far
away from N. Rather, the peers would need to bracket N.</t>
</list></t>
<t>In the first case, no change is needed.</t>
<t>In the second case, N MUST attempt to CONNECT to the new peers
and if it is successful it MUST adjust its neighbor set
accordingly. Note that it can maintain the now inferior peers as
neighbors, but it MUST remember the closer ones.</t>
<t>The third case implies that a neighbor has disappeared, most
likely because it has simply been disconnected but perhaps because
of overlay instability. N MUST PING the questionable peers to
discover if they are indeed missing and if so, remove them from
its neighborhood table.</t>
<t>After any PINGs and CONNECTs are done, if the neighborhood
table changes, the peer sends an UPDATE request to each of its
neighbors that was in either the old table or the new table. These
UPDATE requests are what ends up filling in the
predecessor/successor tables of peers that this peer is a neighbor
to. A peer MUST NOT enter itself in its successor or predecessor
table and instead should leave the entries empty.</t>
<t>A peer N which is responsible for a resource-id R discovers
that the replica set for R (the next two nodes in its successor
set) has changed, it MUST send a STORE for any data associated
with R to any new node in the replica set. It SHOULD not delete
data from peers which have left the replica set.</t>
<t>When a peer N detects that it is no longer in the replica set
for a resource R (i.e., there are three predecessors between N and
R), it SHOULD delete all data associated with R from its local
store.</t>
</section>
<section title="Stabilization">
<t>There are four components to stabilization: <list
style="numbers">
<t>exchange UPDATES will all peers in its routing table to
exchange state</t>
<t>search for better peers to place in its finger table</t>
<t>search to determine if the current finger table size is
sufficiently large</t>
<t>search to determine if the overlay has partitioned and
needs to recover</t>
</list></t>
<t>A peer MUST periodically send an UPDATE request to every peer
in its routing table. The purpose of this is to keep the
predecessor and successor lists up to date and to detect
connection failures. The default time is about every ten minutes,
but the enrollment server SHOULD set this in the configuration
document using the "chord-128-2-16+-update-frequency" element
(denominated in seconds.) A peer SHOULD randomly offset these
UPDATE requests so they do not occur all at once. If an UPDATE
request fails or times out, the peer MUST mark that entry in the
neighbor table invalid and attempt to reestablish a connection. If
no connection can be established, the peer MUST attempt to
establish a new peer as its neighbor and do whatever replica set
adjustments are required.</t>
<t>Periodically a peer should select a random entry i from the
finger table and do a PING to peer (n+2^(numBitsInPeerId-i). The
purpose of this is to find a more accurate finger table entry if
there is one. This is done less frequently than the connectivity
checks in the previous section because forming new connections is
somewhat expensive and the cost needs to be balanced against the
cost of not having the most optimal finger table entries. The
default time is about every hour, but the enrollment server SHOULD
set this in the configuration document using the
"chord-128-2-16+-ping-frequency" element (denominated in seconds).
If this returns a different peer than the one currently in this
entry of the peer table, then a new connection should be formed to
this peer and it should replace the old peer in the finger
table.</t>
<t>As an overlay grows, more than 16 entries may be required in
the finger table for efficient routing. To determine if its finger
table is sufficiently large, one an hour the peer should perform a
PING to determine whether growing its finger table by four entries
would result in it learning at least two peers that it does not
already have in its neighbor table. If so, then the finger table
SHOULD be grown by four entries. Similarly, if the peer observes
that its closest finger table entries are also in its neighbor
table, it MAY shrink its finger table to the minimum size of 16
entries. [[OPEN ISSUE: there are a variety of algorithms to gauge
the population of the overlay and select an appropriate finger
table size. Need to consider which is the best combination of
effectiveness and simplicity.]]</t>
<t>To detect that a partitioning has occurred and to heal the
overlay, a peer P MUST periodically repeat the discovery process
used in the initial join for the overay to locate an appropriate
bootstrap peer, B. If an overlay has multiple mechanisms for
discovery it should randomly select a method to locate a bootstrap
peer. P should then send a PING for its own peer-ID routed through
B. If a response is received from a peer S', which is not P's
successor, then the overlay is partitioned and P should send a
CONNECT to S' routed through B, followed by an UPDATE sent to S'.
(Note that S' may not be in P's neighborhood table once the
overlay is healed, but the connection will allow S' to discover
appropriate neighbor entries for itself via its own
stabilization.)</t>
</section>
</section>
<section title="Leaving">
<t>Peers SHOULD send a LEAVE request prior to exiting the DHT. Any
peer which receives a LEAVE for a peer n in its neighbor set must
remove it from the neighbor set, update its replica sets as
appropriate (including STOREs of data to new members of the replica
set) and send UPDATEs containing its new predecessor and successor
tables.</t>
</section>
</section>
</section>
<section anchor="secEnrollment" title="Enrollment and Bootstrap">
<section title="Discovery">
<t>When a peer first joins a new overlay, it starts with a discovery
process to find an enrollment server. Related work to the approach
used here is described in <xref
target="I-D.garcia-p2psip-dns-sd-bootstrapping"></xref> and <xref
target="I-D.matthews-p2psip-bootstrap-mechanisms"></xref>. The peer
first determines the overlay name. This value is provided by the user
or some other out of band provisioning mechanism. If the name is an IP
address, that is directly used otherwise the peer MUST do a DNS SRV
query using a Service name of "p2p_enroll" and a protocol of tcp to
find an enrollment server.</t>
<t>If the overlay name ends in .local, then the DNS SRV lookup is done
using implement <xref target="I-D.cheshire-dnsext-dns-sd"></xref> with
a Service name of "p2p_menroll" can also be tried to find an
enrollment server. If they implement this, the user name can be used
as the Instance Identifier label.</t>
<t>Once an address for the enrollment servers is determined, the peer
forms an HTTPS connection to that IP address. The certificate MUST
match the overlay name as described in <xref target="RFC2818"></xref>.
The peer then performs a GET to the URL formed by appending a path of
"/p2psip/enroll" to the overlay name. For example, if the overlay name
was example.com, the URL would be
"https://example.com/p2psip/enroll".</t>
<t>The result is an XML configuration file with the syntax described
in the following section.</t>
</section>
<section anchor="sec-configuration" title="Overlay Configuration">
<t>This specification defines a new content type
"application/p2p-overlay+xml" for an MIME entity that contains overlay
information. This information is fetched from the enrollment server,
as described above. An example document is shown below.</t>
<figure>
<artwork><![CDATA[
<?xml version="1.0" encoding="UTF-8"?>
<overlay name="chord.example.com" expiration="86400">
<dht name="chord-128-2-8"/>
<root-cert>[DER certificate here]</root-cert>
<required-usage name="SIP"/>
<credential-server url="https://www.example.com/csr"/>
<bootstrap-peer address="192.0.2.2" port="5678"/>
<bootstrap-peer address="192.0.2.3" port="5678"/>
<bootstrap-peer address="192.0.2.4" port="5678"/>
<multicast-bootstrap="192.0.2.99" port="5678"/>
</overlay>
]]></artwork>
</figure>
<t>The file MUST be a well formed XML document and it SHOULD contain
an encoding declaration in the XML declaration. If the charset
parameter of the MIME content type declaration is present and it is
different from the encoding declaration, the charset parameter takes
precedence. Every application conferment to this specification MUST
accept the UTF-8 character encoding to ensure minimal
interoperability. The namespace for the elements defined in this
specification is urn:ietf:params:xml:ns:p2p:overlay.</t>
<t>The file can contain multiple "overlay" elements where each one
contains the configuration information for a different overlay. Each
"overlay" has the following attributes:</t>
<t><list style="hanging">
<t hangText="name:">name of the overlay</t>
<t hangText="expiration:">time in future at which this overlay
configuration is not longer valid and need to be retrieved again.
This is expressed in seconds from the current time.</t>
</list></t>
<t>Inside each overlay element, the following elements can occur:</t>
<t><list style="hanging">
<t hangText="dht -">This element has an attribute called name that
describes which DHT algorithm is being used.</t>
<t hangText="root-cert -">This element contains a DER encoded
X.509v3 certificate that is the root trust store used to sign all
certificates in this overlay. There can be more than one of
these.</t>
<t hangText="required-usage -">This element has an attribute
called "name" that describes a usage that peers in this overlay
are required to support. More than one required-usage element may
be present.</t>
<t hangText="credential-server -">This element contains the URL at
which the credential server can be reached in a "url" element.
This URL MUST be of type "https:". More than one credential-server
element may be present.</t>
<t hangText="bootstrap-peer -">This elements represents the
address of one of the bootstrap peers. It has an attribute called
"address" that represents the IP address (either IPv4 or IPv6,
since they can be distinguished) and an attribute called "port"
that represents the port. More than one bootstrap-peer element may
be present.</t>
<t hangText="multicast-bootstrap -">This element represents the
address of a multicast address and port that may be used for
bootstrap and that peers SHOULD listen on to enable bootstrap. It
has an attributed called "address" that represents the IP address
and an attribute called "port" that represents the port. More than
one "multicast-bootstrap" element may be present.</t>
<t hangText="iterative-permitted -">This element indicates that
iterative routing (see <xref target="sec-route-query"></xref>) MAY
be used. If iterative routing is permitted, then this value MUST
be set to "TRUE". Otherwise, it SHOULD be absent, but MAY be set
to "FALSE".</t>
</list></t>
<t>[[TODO: Do a RelaxNG grammar.]]</t>
</section>
<section anchor="sec-credentials" title="Credentials">
<t>If the configuration document contains a credential-server element,
credentials are required to use the DHT. A peer which does not yet
have credentials MUST contact the credential server to acquire
them.</t>
<t>In order to acquire credentials, the peer generates an asymmetric
key pair and then generates a "Simple Enrollment Request" (as defined
in <xref target="I-D.ietf-pkix-2797-bis"></xref>) and sends this over
HTTPS as defined in <xref target="I-D.ietf-pkix-cmc-trans"></xref> to
the URL in the credential-server element. The subjectAltName in the
request MUST contain the required user name(s).</t>
<t>The credential server MUST authenticate the request using HTTP
digest <xref target="RFC2617"></xref>. If the authentication succeeds
and the requested user name(s) is acceptable, the server and returns a
certificate. The SubjectAltName field in the certificate contains the
following values:</t>
<t><list style="symbols">
<t>One or more Peer-IDs which MUST be cryptographically random
<xref target="RFC4086"></xref>. These MUST be chosen by the
credential server in such a way that they are unpredictable to the
requesting user.</t>
<t>The names this user is allowed to use in the overlay</t>
</list></t>
<t>The certificate is returned in a "Simple Enrollment Response".</t>
<t>The client MUST check that the certificate returned was signed by
one of the certificates received in the "root-cert" list of the
overlay configuration data. The peer then reads the certificate to
find the Peer-IDs it can use.</t>
<section title="Credentials for HIP">
<t>When RELOAD is used with HIP, the certificates MUST be generated
so that: <list style="symbols">
<t>Each node is assigned a unique ORCHID.</t>
<t>The peer-id can be uniquely determined from the ORCHID.</t>
</list> Because in general, ORCHIDs are shorter than peer-ids,
this means that the ORCHIDS MUST be generated first and MUST be
cryptographically random in order to make the peer-ids
cryptographically random. The mapping function used to produce the
peer-id from the ORCHID MUST be the same as that used by the DHT to
produce resource-ids from Unhashed-IDs.</t>
<t>In addition to the usual attributes, when HIP is in use
certificates MUST contain a subjectAltName with an iPAddress value
containing the HIP ORCHID. This allows these certificates to be used
by the HIP peers during the HIP base exchange.</t>
</section>
</section>
<section title="Locating a Peer">
<t>In order to join the overlay, the peer MUST contact a peer.
Typically this means contacting the bootstrap peers, since they are
guaranteed to have public IP addresses (the system should not
advertise them as bootstrap peers otherwise). If the peer has cached
peers it SHOULD contact them first by sending a PING request to the
known peer address with the destination peer-id set to that peer's
peer-id.</t>
<t>If no cached peers are available, then the peer SHOULD send a PING
request to the address and port found in the broadcast-peers element
in the configuration document. This MAY be a multicast or anycast
address. The PING should use the wildcard peer-id as the destination
peer-id.</t>
<t>The responder peer that receives the PING request SHOULD check that
the overlay name is correct and that the requester peer sending the
request has appropriate credentials for the overlay before responding
to the PING request even if the response is only an error.</t>
<t>When the requester peer finally does receive a response from some
responding peer, it can note the peer-id in the response and use this
peer-id to start sending requests to join the DHT as described in
<xref target="sec-join-leave-maint"></xref> and <xref
target="sec-dht-maint"></xref>.</t>
<t>After a peer has successfully joined the overlay network, it SHOULD
periodically look at any peers to which it has managed to form direct
connections. Some of these peers MAY be added to the cached-peers list
and used in future boots. Peers that are not directly connected MUST
NOT be cached. The RECOMMENDED number of peers to cache is 10.</t>
</section>
</section>
<section title="Usages ">
<section anchor="sec.usage-reqts" title="Generic Usage Requirements">
<t>A new usage MUST specify the following information:</t>
<t><list style="symbols">
<t>The kind-ids which the usage defines and what each kind
means.</t>
<t>The data model for the data being stored (single value, array,
dictionary, etc.) for each kind</t>
<t>Access control rules for each kind, indicating what credentials
are allowed to read and write that kind-id at a given
location.</t>
<t>The minimum amounts of data of each kind that a conformant
implementation MUST store.</t>
</list></t>
<t>While each kind MUST define what data model is used for its data,
that does not mean that it must define new data models. Where
practical, kind SHOULD use the build-in data models. However, they MAY
define any new required data models. The intention is that the basic
data model set be sufficient for most applications/usages.</t>
<t><list style="hanging">
<t hangText="Note:">New usages MAY reuse existing kind-ids. New
kind-ids only need to be defined where different data is stored or
different behavior is required.</t>
</list></t>
</section>
<section title="SIP Usage">
<t>The SIP usage allows a RELOAD overlay to be used as a distributed
SIP registrar/proxy network. The basic function of the SIP usage is to
allow Alice to start with a SIP URI (e.g., "bob@dht.example.com") and
end up with a connection which Bob's SIP UA can use to pass SIP
messages back and forth to Alice's SIP UA. Provides the following
three functions:</t>
<t><list style="symbols">
<t>Mapping SIP URIs that are not GRUUs to the overlay peer
responsible for the SIP UA.</t>
<t>Mapping SIP GRUUs to the DHT peer responsible for the SIP
UA.</t>
<t>Forming a connection directly to a DHT peer that is used to
send SIP messages to the SIP UA.</t>
</list></t>
<t><xref target="sec-sip-usage"></xref> provides an overview of how
these fit together.</t>
<section title="SIP-REGISTRATION kind">
<t>The first mapping is provided using the SIP-REGISTRATION
kind-id:</t>
<t><list style="hanging">
<t hangText="Kind IDs">The Unhashed-ID for the SIP-REGISTRATION
kind-id is a URI, typically the AOR for the user. The data
stored is a sip-registration-data, which can contain either
another URI or a destination list to the peer which is acting
for the user. [[TODO: we want to somehow put caller-prefs in
here along with the route list, but I'm not sure how to do it
yet.]]</t>
<t hangText="Data Model">The data model for the SIP-REGISTRATION
kind-id is dictionary. The dictionary key is the peer-id of the
storing peer. This allows each peer (presumably corresponding to
a single device) to store a single route mapping.</t>
<t hangText="Access Control">If certificate-based access control
is being used, stored data of kind-id SIP-REGISTRATION must be
signed by a certificate which (1) contains user name matching
the storing URI used as the Unhashed-ID for the resource-id and
(2) contains a peer-id matching the storing dictionary key.</t>
<t hangText="Data Sizes">Peers MUST be prepared to store
SIP-REGISTRATION values of up to 10 kilobytes and must be
prepared to store up to 10 values for each user name.</t>
</list></t>
<t>The contents of the SIP-REGISTRATION kind are</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
typedef string sip_registration_uri;
struct {
string contact_prefs;
peer_id destination_list<65000>;
} sip_registration_route;
enum {sip_registration_uri_type(1), sip_registration_route_type(2)} sip_registration_type;
select {
case sip_registration_uri_type:
sip_registration_uri registration_uri;
case sip_registration_route_type:
sip_registration_route registration_route;
} sip_registration_data_;
public struct {
sip_registration_data_ registration_data;
} sip_registration;
]]></artwork>
</figure>
<t>A registration may contain either a URI (type code 0x01) or a
contact preferences structure and a destination list (type code
0x01). The leading byte indicates the type.</t>
</section>
<section title="GRUUs">
<t>GRUUs do not require storing data in the DHT. Rather, they are
constructed by embedding a base64-encoded destination list in the gr
URI parameter of the GRUU. The base64 encoding is done with the
alphabet specified in table 1 of RFC 4648 with the exception that ~
is used in place of =. An example GRUU is
"sip:alice@example.com;gr=MDEyMzQ1Njc4OTAxMjM0NTY3ODk~". When a peer
needs to route a message to a GRUU in the same P2P network, it
simply uses the destination list and connects to that peer.</t>
<t>Anonymous GRUUs are done in roughly the same way but require
either that the enrollment server issue a different peer-id for each
anonymous GRUU required or that a destination list be used that
includes a peer that compresses the destination list to stop the
peer-id from being revealed.</t>
</section>
<section title="SIP Connect">
<t>Once the destination list for a user has been identified, the
calling peer uses the CONNECT request to form a connection to the
peer identified by the destination list. The CONNECT request MUST
contain the connect-application value of 5160 (SIP). If
certificate-based authentication is in use, the responding peer MUST
present a certificate with a peer-id matching the terminal entry in
the route list.</t>
<t>[[TODO: Note that this constrains destination lists from hiding
the last peer-id when used here. I think that's OK, but we should
take a look]]</t>
<t>Once the association has been formed, the calling peer sends
generic SIP messages down the new association and ordinary SIP
procedures are followed.</t>
</section>
<section title="SIP Tunnel">
<t>This usage allows two peers to exchange SIP messages across the
overlay using the TUNNEL method. TUNNEL is provided as an
alternative to using CONNECT because it allows a SIP message to be
sent immediately, without the delay associated with CONNECT. For a
simple SIP exchange, it may result in fewer messages being sent.</t>
<t>An implementation SHOULD use CONNECT for a dialog that is
expected to endure for sufficient time and exchange significant
numbers of messages. An implementation MAY establish an initial
dialog using TUNNELing and then migrate it to a direct dialog opened
with CONNECT once that negotiation is complete.</t>
<t>As an application of TUNNEL, this usage defines the following
items:</t>
<t><list style="symbols">
<t>For SIP, the application attribute is 5060.</t>
<t>The application MAY establish any dialog using TUNNEL if it
expects to replace it once a CONNECT request completes. The
application SHOULD NOT exchange messages with another SIP UA
repeatedly using a TUNNEL unless it is unable to complete a
CONNECT.</t>
<t>The Replaces header should be used to migrate dialogs
established via TUNNEL to a direct connection.</t>
<t>The dialogid is the GRUU of the destination of the
request.</t>
<t>By using the GRUU of the destination as the dialogid, the
receiving peer is able to deliver the message to the appropriate
process without parsing the SIP message.</t>
</list></t>
<t>In constructing the message, the SIP UA forms the message as if
it were being routed directly to the GRUU of the destination. The
SIP stack hands the message to RELOAD for delivery. Although the
message is passed through a sequence of untrusted peers, it is not
subject to modification by those peers because of the message's
signature.</t>
<t>OPEN ISSUE: should specify how to request encryption of the
message end-to-end.</t>
<t><list style="hanging">
<t hangText="Note:">The easiest implementation of TUNNEL is
likely to default to sending all messages across a TUNNEL when
the first message is sent to a new destination GRUU and
simultaneously issuing a CONNECT. Messages then continue through
the TUNNEL until the CONNECT completes, at which point they are
delivered via the new connection.</t>
</list></t>
<t>OPEN ISSUE: If the tunneling vs direct decision can be made
equivalently to a link-layer decision, it may not be necessary to
modify the dialog or inform the SIP UA in any way that it has now
obtained a direct route.</t>
</section>
</section>
<section anchor="sec-stun-turn-usage" title="TURN Usage">
<t>When a node starts up, it joins the overlay network and forms
several connection in the process. If the ICE stage in any of these
connection return a reflexive address that is not the same as the
peers perceived address, then the peers is behind a NAT and not an
candidate for a TURN server. Additionally, if the peers IP address is
in the private address space range, then it is not a candidate for a
TURN server. Otherwise, the peer SHOULD assume it is a potential TURN
server and follow the procedures below.</t>
<t>If the node is a candidate for a TURN server it will insert some
pointers in the overlay so that other peers can find it. The overlay
configuration file specifies a turnDensity parameter that indicates
how many times each TURN server should record itself in the overlay.
Typically this should be set to the reciprocal of the estimate of what
percentage of peers will act as TURN servers. For each value, called
d, between 1 and turnDensity, the peer forms a Unhashed-ID by
concatenating its peer-ID and the value d. This Unhashed-ID is hashed
to form a Resource-ID. The address of the peer is stored at that
Resource-ID using type TURN-SERVICE and the turn-server
production:</t>
<t><list style="hanging">
<t hangText="Note:">Correct functioning of this algorithm depends
critically on having turnDensity be an accurate estimate of the
true density of TURN servers. If turnDensity is too high, then the
process of finding TURN servers becomes extremely expensive as
multiple candidate resource-ids must be probed.</t>
</list></t>
<t>Peers peers that provide the STUN-Relay server type need to support
the TURN extensions to STUN for media relay of both UDP and TCP
traffic as defined in <xref target="I-D.ietf-behave-turn"></xref> and
<xref target="I-D.ietf-behave-tcp"></xref>.</t>
<figure>
<!--begin-pdu-->
<artwork><![CDATA[
public struct {
uint8 iteration;
ip_address_and_port address;
} turn_server;
]]></artwork>
<!--
turn-server = ip-address-and-port
ip-address-and-port = ip-address
ip-port
ip-address = ip-address-type
ip4-address / ip6-address
ip-address-type = ip4-address-type / ip6-address-type ; uint8
ip4-address-type = 0x01
ip6-address-type = 0x02
ip4-address = uint32
ip6-address = uint128
ip-port = uint16
iteration = uint8
-->
</figure>
<t>[[OPEN ISSUE: This structure only works for TURN servers that have
public addresses. It may be possible to use TURN servers that are
behind well-behaved NATs by first ICE connecting to them. If we decide
we want to enable that, this structure will need to change to either
be a peer-id or include that as an option.]]</t>
<t><list style="hanging">
<t hangText="Kind IDs">This usage defines the TURN-SERVICE kind-id
to indicate that a peer is willing to act as a TURN server. The
FIND command MUST return results for the TURN-SERVICE kind-id.</t>
<t hangText="Data Model">The TURN-SERVICE stores a single value
for each resource-id.</t>
<t hangText="Access Control">If certificate-based access control
is being used, stored data of kind TURN-SERVICE MUST be
authenticated by a certificate which contains a peer-id which when
hashed with the iteration counter produces the resource-id being
stored at.</t>
<t hangText="Data Sizes">TURN-SERVICE values are of fixed size.
Peers MUST be prepared to store values with iteration counter of
up to 100.</t>
</list></t>
<t>The data is stored in a data structure with the IP address of the
server and an indication whether the address is an IPv4 or IPv6
address. The Unhashed-ID used to form the storage Resource-ID is
simply the peer-id. The access control rule is that the certificate
used to sign the request must contain a peer-id that when hashed would
match the Resource-ID where the data is being stored.</t>
<t>Peers can find other servers by selecting a random Resource-ID and
then doing a FIND request for the appropriate server type with that
Resource-ID. The FIND request gets routed to a random peer based on
the Resource-ID. If that peer knows of any servers, they will be
returned. The returned response may be empty if the peer does not know
of any servers, in which case the process gets repeated with some
other random Resource-ID. As long as the ratio of servers relative to
peers is not too low, this approach will result in finding a server
relatively quickly.</t>
<t>Open issues: Should there be low and high bandwidth version of
STUN-Relay one can find? Low would be usable for signaling type things
and high would be usable for audio, video, and others.</t>
</section>
<section title="Certificate Store Usage">
<t>The Certificate Store usage allows a peer to store its certificate
in the overlay, thus avoiding the need to send a certificate in each
message - a reference may be sent instead.</t>
<t>A user/peer MUST store its certificate at resource-ids derived from
two Unhashed-IDs:</t>
<t><list style="symbols">
<t>The user names in the certificate.</t>
<t>The peer-ids in the certificate.</t>
</list></t>
<t>Note that in the second case the certificate is not stored at the
peer's peer-id but rather at a hash of the peer's peer-id. The
intention here (as is common throughout RELOAD) is to avoid making a
peer responsible for its own data.</t>
<t>A peer MUST ensure that the user's certificates are stored in the
DHT when joining and redo the check about every 24 hours after that.
Certificate data should be stored with an expiry time of 60 days. When
a client is checking the existence of data, if the expiry is less than
30 days, it should be refreshed to have an expiry of 60 days. The
certificate information is frequently used for many operations, and
peers should cache it for 8 hours.</t>
<t><list style="hanging">
<t hangText="Kind IDs">This usage defines the CERTIFICATE kind-id
to store a peer or user's certificate.</t>
<t hangText="Data Model">The data model for CERTIFICATE data is
array.</t>
<t hangText="Access Control">The CERTIFICATE MUST contain a
peer-id or user name which, when hashed, maps the resource-id at
which the value is being stored.</t>
<t hangText="Data Sizes">Peers MUST be prepared to store at least
10 certificates of sizes up to 1K each.</t>
</list></t>
</section>
<section title="HIP Tunnel">
<t>This usage allows two peers to exchange HIP messages across the
overlay using the TUNNEL method. This is meant to be used as part of
the HIP BONE architecture described in <xref
target="I-D.camarillo-hip-bone"></xref>.</t>
<t>As an application of TUNNEL, this usage defines the following
items:</t>
<t><list style="symbols">
<t>For HIP, the application attribute is TBD (IANA port
number).</t>
<t>The dialogid is empty and set to zero length.</t>
</list></t>
<t>In order to route HIP messages correctly, there needs to be an
unambiguous mapping between the ORCHID assigned to each HIP node and
the peer-id assigned to that node. The ORCHID MUST be used as the
Unhashed-ID to generate the peer-id. [TODO: We need a general scheme
for mapping Unhashed-IDs to IDs so they don't collide. This isn't the
place to define it.] Messages to a given ORCHID are then routed to the
mapped peer-id. <xref target="sec-credentials"></xref> describes more
considerations for the generation of ORCHIDs for use with RELOAD.</t>
<t>TODO - should discuss interaction of HIP and P2P retransmission
timers</t>
</section>
<section title="Diagnostic Usage" toc="default">
<t>[[TODO: reduce text of motivation description in the next
version]]</t>
<t>The development and deployment of a peer-to-peer system is a
continuous process. The developers write code which is tested on a
scale that may be smaller than the actual deployment size. After this
local testing, the code is deployed in a real environment. Bugs arise
during development and deployment phases. The designers of the
peer-to-peer system need mechanisms which can help identify problems
and bugs in a peer-to-peer system during development and deployment
phases. Peer-to-peer systems are an example of a distributed system
and it is not a trivial task to provide protocol mechanisms, tools and
techniques to identify problems that may arise in such systems.</t>
<t>The diagnostic mechanisms can broadly be classified into online and
offline mechanisms. The online mechanisms attempt to identify faults
in a running system where as offline mechanisms try to infer faults by
gathering the log files of machines participating in a distributed
system.</t>
<t>In a peer-to-peer system, a peer maintains routing state to forward
messages according to the overlay protocol being used. In addition, a
peer stores information published by other peers. The routing and
storage of resources consume network, space (memory), and CPU
resources. A peer also needs to keep track of how long the P2PSIP
application has been running and the last time peers in the routing
table were last contacted. During development and deployment phase, an
overlay designer needs mechanisms to query some or all of the above
mentioned information.</t>
<t>The overlay designer may also treat overlay as a black box and
determine if the routing mechanisms are working correctly under
various levels of churn.</t>
<t>Thus, there are at least two types of online diagnostic mechanisms:
1) state acquisition 2) black-box diagnostics</t>
<section title="State Acquisition Mechanisms">
<t>The protocol provides a DIAGNOSTIC method [TODO] which queries
the peer for its routing state, average bandwidth, CPU utilization,
and storage state. The DIAGNOSTIC request should typically be sent
over a reliable transport protocol as the response will likely
exceed UDP MTU size. The state acquisition mechanism can be used to
construct a local view of the connectivity state of the system. It
can also be used to construct a geographical map of the system.</t>
<t>Below, we identify potential issues with the state acquisition
mechanisms.</t>
<t>Security: If any peer can query the routing or storage state of
any other peer, then clearly privacy and security concerns arise. To
address this, the state acquisition mechanisms need an access list
like mechanism so that only the overlay implementer can query the
state of all the nodes. Alternatively, the state acquisition
mechanisms are only enabled during the development phase or are only
enabled for 'admin' users.</t>
<t>Scalability: It is possible to query the state of few hundred or
a few thousand nodes (as it is currently done in our live system on
Planet lab); however, a serial state acquisition of a million node
is a non starter. In large scale networks, one option is to query
the state of few hundred nodes and to construct an high level
connectivity map. CAIDA [ref] collects data at a few vantage points
to construct BGP maps.</t>
<t>Instantaneous vs. long term state: Another issue with these state
acquisitions mechanisms is whether they acquire the instantaneous
state snapshot or an exponential moving average or a list of
snapshots over a period of time. For diagnostic metrics such as CPU
utilization, an exponential moving average metric is also helpful in
addition to the instantaneous snapshot.</t>
<t>Pull vs. push: The state acquisition mechanisms can either be
pull-based or push-based or a combination of both. In pull-based
mechanisms, peer explicitly request state of another peer. This may
not be sufficient because pull-based mechanisms require a to
periodically poll a peer for any change state. In a push-based
mechanism, peers advertise any change in certain metrics to their
routing or neighbor peers. As an example of push-based mechanism, a
peer which starts to relay a call may indicate a change in its
bandwidth to its routing or neighbor peers in a PING message.</t>
<t>Development vs. deployment: A hard problem is to decide which
diagnostics are absolutely necessary during deployment and which are
needed during development.</t>
<t>Clearly, complete state acquisition has security concerns in a
deployed system. The other option an overlay implementer can use is
to run a few peers and have complete control over the functionality
of these peers. These peers are same as other peers with the
difference that an overlay implementer can explicitly query the
state of these peers. It can then use this information to 'crawl'
the overlay network and construct a local map of the network.</t>
</section>
<section title="Black-box diagnostics">
<t>[[TODO: a better name for this section]]</t>
<t>Black-box diagnostics: DHTs are examples of structured
peer-to-peer networks and they allow nodes to store key/value pairs
in the overlay. A simple diagnostic mechanism is to treat the
overlay as a black-box: publish several key/value pairs at one peer
and then look them up from another peer. For this kind of diagnostic
mechanism, clients are more suitable as they do not provide any
routing or storage services to the overlay and can connect to an
arbitrary peer.</t>
</section>
<t>The Diagnostic Usage allow a peer to report various statistics
about itself that may be useful for diagnostics or performance
management. It can be used to discover information such as the
software version, uptime, and performance statistics of a peer. The
usage defines several new kinds which can be retrieved to get the
statistics. The peer-id is directly used when retrieving data so no
Unhashed-ID is defined. The access control model for all of these is
local policy defined by the peer. The peer MAY have a list of users
(such as "admin") that it is willing to return the information for and
restrict access to users with that name. The access control can be
determined on a per kind basis - for example, a node may be willing to
return the software version to any users while specific information
about performance may not be returned.</t>
<t>The following kinds are defined:</t>
<t><list style="hanging">
<t hangText="SOFTWARE_VERSION">A single value element containing a
US-ASCII string that identifies the manufacture, model, and
version of the software.</t>
<t hangText="UPTIME">A single value element containing an unsigned
64-bit integer specifying the time the nodes has been up in
seconds.</t>
<t hangText="AS_NUMBER">A single value element containing the
Autonomous System [TODO REF] number as an unsigned 32-bit integer.
Zero is returned if the AS number is unknown.</t>
<t>(OPEN ISSUES: How to determine a AS number? This metric is
primarily used for advertising and locating STUN/TURN servers. A
TURN server is inserted and looked up under H(AS). What if there
are no TURN servers in the same AS? )</t>
<t hangText="CPU_UTILIZATION">A single value element containing an
unsigned 8-bit integer representing the percentage CPU load from 1
to 100.</t>
<t>(OPEN ISSUE: It is not a very precise metric.)</t>
<t hangText="DATA_STORED">A single value element containing an
unsigned 64-bit integer representing the number of bytes of data
being stored by this node.</t>
<t hangText="MESSAGES_SENT">An array element containing the number
of messages sent and received. The array is indexed by method
code. Each entry in the array is a pair of unsigned 64-bit
integers (packed end to end) representing sent and received.</t>
<t hangText="INSTANCES_STORED">An array element containing the
number of instances of each kind stored. The array is index by
kind-id. Each entry is an unsigned 64-bit integer.</t>
<t hangText="ROUTING_TABLE_SIZE">A single value element containing
an unsigned 32-bit integer representing the number of peers in the
node's routing table.</t>
<t hangText="NEIGHBOR_TABLE_SIZE">A single value element
containing an unsigned 32-bit integer representing the number of
peers in the node's neighbor table.</t>
<t hangText="EWMA_BYTES_SENT">A single value element containing an
unsigned 32-bit integer representing an exponential weighted
average of bytes sent by this peer.</t>
<t hangText="EWMA_BYTES_RCVD">A single value element containing an
unsigned 32-bit integer representing an exponential weighted
average of bytes received by this peer.</t>
<t hangText="LAST_CONTACT">A single value element containing an
unsigned 32-bit integer specifying the time in number of seconds
the node was last contacted.</t>
<t hangText="MEMORY_FOOTPRINT">A single value element containing
an unsigned 32-bit integer representing the memory footprint of
the peer program in kilo bytes.</t>
<t hangText="RTT">A single value element containing an unsigned
32-bit integer specifying the recent RTT estimate in ms between
two peers.</t>
</list></t>
<t>[[TODO: We would like some sort of bandwidth measurement, but we're
kind of unclear on the units and representation.]]</t>
<section title="Diagnostic Metrics for a P2PSIP Deployment">
<t>Clearly, all diagnostic metrics are useful during development and
testing. The hard question is which metrics are absolutely necessary
for a deployed P2PSIP system. We attempt to identify these metrics
and classify them under 'resource' and 'peer' metrics.</t>
<t>For 'resource' metric, we identify CPU_UTILIZATION,
EWMA_BYTES_SENT, EWMA_BYTES_RCVD, and MEMORY_FOOTPRINT as the key
metrics and for 'peer' metric we identify UPTIME, LAST_CONTACT, and
RTT as the metrics that are crucial for a deployed P2PSIP
system.</t>
<t>(OPEN QUESTION: any other metrics?)</t>
<t>(OPEN: Below, we sketch how these metrics can be used. A peer can
use EWMA_BYTES_SENT and EWMA_BYTES_RCVD of another peer to infer
whether it is acting as a media relay. It may then choose not to
forward any requests for media relay to this peer. Similarly, among
the various candidates for filling up routing table, a peer may
prefer a peer with a large UPTIME value, small RTT, and small
LAST_CONTACT value. )</t>
</section>
<!-- <t>[[TODO: Why not MIB]]</t> -->
</section>
</section>
<section title="Security Considerations">
<section title="Overview">
<t>RELOAD provides a generic storage service, albeit one designed to
be useful for P2P SIP. In this section we discuss security issues that
are likely to be relevant to any usage of RELOAD. In <xref
target="section.sip-issues"></xref> we describe issues that are
specific to SIP.</t>
<t>In any DHT, any given user depends on a number of peers with which
they have no well-defined relationship except that they are fellow
members of the DHT. In practice, these other nodes may be friendly,
lazy, curious, or outright malicious. No security system can provide
complete protection in an environment where most nodes are malicious.
The goal of security in RELOAD is to provide strong security
guarantees of some properties even in the face of a large number of
malicious nodes and to allow the DHT to function correctly in the face
of a modest number of malicious nodes.</t>
<t>P2PSIP deployments require the ability to authenticate both peers
and resources (users) without the active presence of a trusted entity
in the system. We describe two mechanisms. The first mechanism is
based on public key certificates and is suitable for general
deployments. The second is based on an overlay-wide shared symmetric
key and is suitable only for limited deployments in which the
relationship between admitted peers is not adversarial.</t>
</section>
<section title="Attacks on P2P Overlays">
<t>The two basic functions provided by DHT nodes are storage and
routing: some node is responsible for storing a peer's data and for
allowing a peer to fetch other peer's data. Some other set of nodes
are responsible for routing messages to and from the storing nodes.
Each of these issues is covered in the following sections.</t>
<t>P2P overlays are subject to attacks by subversive nodes that may
attempt to disrupt routing, corrupt or remove user registrations, or
eavesdrop on signaling. The certificate-based security algorithms we
describe in this draft are intended to protect DHT routing and user
registration information in RELOAD messages.</t>
<t>To protect the signaling from attackers pretending to be valid
peers (or peers other than themselves), the first requirement is to
ensure that all messages are received from authorized members of the
overlay. For this reason, RELOAD transports all messages over DTLS or
TLS, which provides message integrity and authentication of the
directly communicating peer. In addition, when the certificate-based
security system is used, messages and data are digitally signed with
the sender's private key, providing end-to-end security for
communications.</t>
</section>
<section title="Certificate-based Security">
<t>This specification stores users' registrations and possibly other
data in a Distributed Hash table (DHT). This requires a solution to
securing this data as well as securing, as well as possible, the
routing in the DHT. Both types of security are based on requiring that
every entity in the system (whether user or peer) authenticate
cryptographically using an asymmetric key pair tied to a
certificate.</t>
<t>When a user enrolls in the DHT, they request or are assigned a
unique name, such as "alice@dht.example.net". These names are unique
and are meant to be chosen and used by humans much like a SIP Address
of Record (AOR) or an email address. The user is also assigned one or
more peer-IDs by the central enrollment authority. Both the name and
the peer ID are placed in the certificate, along with the user's
public key.</t>
<t>Each certificate enables an entity to act in two sorts of
roles:</t>
<t><list style="symbols">
<t>As a user, storing data at specific Resource-IDs in the DHT
corresponding to the user name.</t>
<t>As a DHT peer with the peer ID(s) listed in the
certificate.</t>
</list></t>
<t>Note that since only users of this DHT need to validate a
certificate, this usage does not require a global PKI. It does,
however, require a central enrollment authority which acts as the
certificate authority for the DHT. This authority signs each peer's
certificate. Because each peer possesses the CA's certificate (which
they receive on enrollment) they can verify the certificates of the
other entities in the overlay without further communication. Because
the certificates contain the user/peer's public key, communications
from the user/peer can be verified in turn.</t>
<t>In order to protect data storage, in the certificate-based security
scheme, all stored data is signed by the owner of the data. This
allows the storing peer to verify that the storer is authorized to
perform a store at that resource-id and also allows any consumer of
the data to verify the provenance and integrity of the data when it
retrieves it.</t>
<t>All implementations MUST implement certificate-based security.</t>
</section>
<section title="Shared-Secret Security">
<t>For small environments where deployment of the PKI necessary to use
a certificate-based model is impractical, RELOAD supports a shared
secret security that relies on a single key that is shared among all
members of the overlay. It is appropriate for small groups that wish
to form a private network without complexity. In shared secret mode,
all the peers share a single symmetric key which is used to key
TLS-PSK <xref target="RFC4279"></xref> or TLS-SRP <xref
target="I-D.ietf-tls-srp"></xref> mode. A peer which does not know the
key cannot form TLS connections with any other peer and therefore
cannot join the overlay.</t>
<t>The shared-secret scheme prohibits unauthorized peers from joining
the overlay, but it provides no protection from a compromised peer
inserting arbitrary resource registrations, performing a Sybil
attack<xref target="Sybil"></xref>, or performing other attacks on the
resources or routing. Thus, it is only safe to use in limited settings
in which peers are not adversarial. In addition, because the messages
and data are not authenticated, each intermediate peer MUST take care
to use TLS and check the other peer's knowledge of the shared secret,
or message insertion is possible.</t>
<t>If the shared secret key for the shared-key security scheme is
discovered by an attacker, then most of the security of the scheme is
lost: an attacker can impersonate any peer to any other peer. Thus,
the shared-secret scheme is only appropriate for small deployments,
such as a small office or ad hoc overlay set up among participants in
a meeting.</t>
<t>One natural approach to a shared-secret scheme is to use a
user-entered password as the key. The difficulty with this is that in
TLS-PSK mode, such keys are very susceptible to dictionary attacks. If
passwords are used as the source of shared-keys, then TLS-SRP is a
superior choice because it is not subject to dictionary attacks.</t>
</section>
<section title="Storage Security">
<t>When certificate-based security is used in RELOAD, any given
Resource-ID/kind-id pair (a slot) is bound to some small set of
certificates. In order to write data in a slot, the writer must prove
possession of the private key for one of those certificates. Moreover,
all data is stored signed by the certificate which authorized its
storage. This set of rules makes questions of authorization and data
integrity - which have historically been thorny for DHTs - relatively
simple.</t>
<t>When shared-secret security is used, then all peers trust all other
peers, provided that they have demonstrated that they have the
credentials to join the overlay at all. The following text therefore
applies only to certificate-based security.</t>
<section title="Authorization">
<t>When a client wants to store some value in a slot, it first
digitally signs the value with its own private key. It then sends a
STORE request that contains both the value and the signature towards
the storing peer (which is defined by the Unhashed-ID construction
algorithm for that particular kind of value).</t>
<t>When the storing peer receives the request, it must determine
whether the storing client is authorized to store in this slot. In
order to do so, it executes the Unhashed-ID construction algorithm
for the specified kind based on the user's certificate information.
It then computes the Resource-ID from the Unhashed-ID and verifies
that it matches the slot which the user is requesting to write to.
If it does, the user is authorized to write to this slot, pending
quota checks as described in the next section.</t>
<t>For example, consider the certificate with the following
properties:</t>
<figure>
<artwork><![CDATA[
User name: alice@dht.example.com
Peer-Id: 013456789abcdef
Serial: 1234
]]></artwork>
</figure>
<t>If Alice wishes to STORE a value of the "SIP Location" kind, the
Unhashed-ID will be the SIP AOR "sip:alice@dht.example.com". The
Resource-ID will be determined by hashing the Unhashed-ID. When a
peer receives a request to store a record at Resource-ID X, it takes
the signing certificate and recomputes the Unhashed-ID, in this case
"alice@dht.example.com". If H("alice@dht.example.com")=X then the
STORE is authorized. Otherwise it is not. Note that the Unhashed-ID
construction algorithm may be different for other kinds.</t>
</section>
<section title="Distributed Quota">
<t>Being a peer in a DHT carries with it the responsibility to store
data for a given region of the DHT. However, if clients were allowed
to store unlimited amounts of data, this would create unacceptable
burdens on peers, as well as enabling trivial denial of service
attacks. RELOAD addresses this issue by requiring each usage to
define maximum sizes for each kind of stored data. Attempts to store
values exceeding this size MUST be rejected (if peers are
inconsistent about this, then strange artifacts will happen when the
zone of responsibility shifts and a different peer becomes
responsible for overlarge data). Because each slot is bound to a
small set of certificates, these size restrictions also create a
distributed quota mechanism, with the quotas administered by the
central enrollment server.</t>
<t>Allowing different kinds of data to have different size
restrictions allows new usages the flexibility to define limits that
fit their needs without requiring all usages to have expansive
limits.</t>
</section>
<section title="Correctness">
<t>Because each stored value is signed, it is trivial for any
retrieving peer to verify the integrity of the stored value. Some
more care needs to be taken to prevent version rollback attacks.
Rollback attacks on storage are prevented by the use of store times
and lifetime values in each store. A lifetime represents the latest
time at which the data is valid and thus limits (though does not
completely prevent) the ability of the storing node to perform a
rollback attack on retrievers. In order to prevent a rollback attack
at the time of the STORE request, we require that storage times be
monotonically increasing. Storing peers MUST reject STORE requests
with storage times smaller than or equal to those they are currently
storing. In addition, a fetching node which receives a data value
with a storage time older than the result of the previous fetch
knows a rollback has occurred.</t>
</section>
<section title="Residual Attacks">
<t>The mechanisms described here provide a high degree of security,
but some attacks remain possible. Most simply, it is possible for
storing nodes to refuse to store a value (i.e., reject any request).
In addition, a storing node can deny knowledge of values which it
previously accepted. To some extent these attacks can be ameliorated
by attempting to store to/retrieve from replicas, but a retrieving
client does not know whether it should try this or not, since there
is a cost to doing so.</t>
<t>Although the certificate-based authentication scheme prevents a
single peer from being able to forge data owned by other peers.
Furthermore, although a subversive peer can refuse to return data
resources for which it is responsible it cannot return forged data
because it cannot provide authentication for such registrations.
Therefore parallel searches for redundant registrations can mitigate
most of the affects of a compromised peer. The ultimate reliability
of such an overlay is a statistical question based on the
replication factor and the percentage of compromised peers.</t>
<t>In addition, when a kind is is multivalued (e.g., an array data
model), the storing node can return only some subset of the values,
thus biasing its responses. This can be countered by using single
values rather than sets, but that makes coordination between
multiple storing agents much more difficult. This is a tradeoff that
must be made when designing any usage.</t>
</section>
</section>
<section title="Routing Security">
<t>Because the storage security system guarantees (within limits) the
integrity of the stored data, routing security focuses on stopping the
attacker from performing a DOS attack on the system by misrouting
requests in the DHT. There are a few obvious observations to make
about this. First, it is easy to ensure that an attacker is at least a
valid peer in the DHT. Second, this is a DOS attack only. Third, if a
large percentage of the peers on the DHT are controlled by the
attacker, it is probably impossible to perfectly secure against
this.</t>
<section title="Background">
<t>In general, attacks on DHT routing are mounted by the attacker
arranging to route traffic through or two nodes it controls. In the
Eclipse attack <xref target="Eclipse"></xref> the attacker tampers
with messages to and from nodes for which it is on-path with respect
to a given victim node. This allows it to pretend to be all the
nodes that are reachable through it. In the Sybil attack <xref
target="Sybil"></xref>, the attacker registers a large number of
nodes and is therefore able to capture a large amount of the traffic
through the DHT.</t>
<t>Both the Eclipse and Sybil attacks require the attacker to be
able to exercise control over her peer IDs. The Sybil attack
requires the creation of a large number of peers. The Eclipse attack
requires that the attacker be able to impersonate specific peers. In
both cases, these attacks are limited by the use of centralized,
certificate-based admission control.</t>
</section>
<section title="Admissions Control">
<t>Admission to an RELOAD DHT is controlled by requiring that each
peer have a certificate containing its peer ID. The requirement to
have a certificate is enforced by using TLS mutual authentication on
each connection. Thus, whenever a peer connects to another peer,
each side automatically checks that the other has a suitable
certificate. These peer IDs are randomly assigned by the central
enrollment server. This has two benefits:</t>
<t><list style="symbols">
<t>It allows the enrollment server to limit the number of peer
IDs issued to any individual user.</t>
<t>It prevents the attacker from choosing specific peer IDs.</t>
</list></t>
<t>The first property allows protection against Sybil attacks
(provided the enrollment server uses strict rate limiting policies).
The second property deters but does not completely prevent Eclipse
attacks. Because an Eclipse attacker must impersonate peers on the
other side of the attacker, he must have a certificate for suitable
peer IDs, which requires him to repeatedly query the enrollment
server for new certificates which only will match by chance. From
the attacker's perspective, the difficulty is that if he only has a
small number of certificates the region of the DHT he is
impersonating appears to be very sparsely populated by comparison to
the victim's local region.</t>
</section>
<section title="Peer Identification and Authentication">
<t>In general, whenever a peer engages in DHT activity that might
affect the routing table it must establish its identity. This
happens in two ways. First, whenever a peer establishes a direct
connection to another peer it authenticates via TLS mutual
authentication. All messages between peers are sent over this
protected channel and therefore the peers can verify the data origin
of the last hop peer for requests and responses without further
cryptography.</t>
<t>In some situations, however, it is desirable to be able to
establish the identity of a peer with whom one is not directly
connected. The most natural case is when a peer UPDATEs its state.
At this point, other peers may need to update their view of the DHT
structure, but they need to verify that the UPDATE message came from
the actual peer rather than from an attacker. To prevent this, all
DHT routing messages are signed by the peer that generated them.</t>
<t>[TODO: this allows for replay attacks on requests. There are two
basic defenses here. The first is global clocks and loose
anti-replay. The second is to refuse to take any action unless you
verify the data with the relevant node. This issue is
undecided.]</t>
<t>[TODO: I think we are probably going to end up with generic
signatures or at least optional signatures on all DHT messages.]</t>
</section>
<section title="Protecting the Signaling">
<t>The goal here is to stop an attacker from knowing who is
signaling what to whom. An attacker being able to observe the
activities of a specific individual is unlikely given the
randomization of IDs and routing based on the present peers
discussed above. Furthermore, because messages can be routed using
only the header information, the actual body of the RELOAD message
can be encrypted during transmission.</t>
<t>There are two lines of defense here. The first is the use of TLS
or DTLS for each communications link between peers. This provides
protection against attackers who are not members of the overlay. The
second line of defense, if certificate-based security is used, is to
digitally sign each message. This prevents adversarial peers from
modifying messages in flight, even if they are on the routing
path.</t>
</section>
<section title="Residual Attacks">
<t>The routing security mechanisms in RELOAD are designed to contain
rather than eliminate attacks on routing. It is still possible for
an attacker to mount a variety of attacks. In particular, if an
attacker is able to take up a position on the DHT routing between A
and B it can make it appear as if B does not exist or is
disconnected. It can also advertise false network metrics in attempt
to reroute traffic. However, these are primarily DoS attacks.</t>
<t>The certificate-based security scheme secures the namespace, but
if an individual peer is compromised or if an attacker obtains a
certificate from the CA, then a number of subversive peers can still
appear in the overlay. While these peers cannot falsify responses to
resource queries, they can respond with error messages, effecting a
DoS attack on the resource registration. They can also subvert
routing to other compromised peers. To defend against such attacks,
a resource search must still consist of parallel searches for
replicated registrations.</t>
</section>
</section>
<section anchor="section.sip-issues" title="SIP-Specific Issues">
<section title="Fork Explosion">
<t>Because SIP includes a forking capability (the ability to
retarget to multiple recipients), fork bombs are a potential DoS
concern. However, in the SIP usage of RELOAD, fork bombs are a much
lower concern because the calling party is involved in each
retargeting event and can therefore directly measure the number of
forks and throttle at some reasonable number.</t>
</section>
<section title="Malicious Retargeting">
<t>Another potential DoS attack is for the owner of an attractive
number to retarget all calls to some victim. This attack is
difficult to ameliorate without requiring the target of a SIP
registration to authorize all stores. The overhead of that
requirement would be excessive and in addition there are good use
cases for retargeting to a peer without there explicit
cooperation.</t>
</section>
<section title="Privacy Issues">
<t>All RELOAD SIP registration data is public. Methods of providing
location and identity privacy are still being studied.</t>
</section>
</section>
</section>
<!--
<section title="Operations and Management Considerations"></section>
-->
<section title="IANA Considerations">
<t>This section contains the new code points registered by this
document. The IANA policies are TBD.</t>
<section title="Overlay Algorithm Types">
<t>IANA SHALL create/(has created) a "RELOAD Overlay Algorithm Type"
Registry. Entries in this registry are strings denoting the names of
DHT algorithms. The registration policy for this registry is TBD.</t>
<t>The initial contents of this registry are:</t>
<t><list style="hanging">
<t hangText="chord-128-2-8 -">The algorithm defined in <xref
target="sec.chord-algorithm"></xref> of this document.</t>
</list></t>
</section>
<section title="Data Kind-Id">
<t>IANA SHALL create/(has created) a "RELOAD Data Kind-Id" Registry.
Entries in this registry are 32-bit integers denoting data kinds, as
described in <xref target="sec.usage-reqts"></xref>. The registration
policy for this registry is TBD.</t>
<t>The initial contents of this registry are:</t>
<t></t>
<texttable>
<ttcol align="left">Kind</ttcol>
<ttcol align="right">Kind-Id</ttcol>
<c>SIP-REGISTRATION</c>
<c>TBD</c>
<c>TURN_SERVICE</c>
<c>TBD</c>
<c>CERTIFICATE</c>
<c>TBD</c>
<c>SOFTWARE_VERSION</c>
<c>TBD</c>
<c>UPTIME</c>
<c>TBD</c>
<c>AS_NUMBER</c>
<c>TBD</c>
<c>CPU_UTILIZATION</c>
<c>TBD</c>
<c>DATA_STORED</c>
<c>TBD</c>
<c>MESSAGES_SENT</c>
<c>TBD</c>
<c>INSTANCES_STORED</c>
<c>TBD</c>
<c>ROUTING_TABLE_SIZE</c>
<c>TBD</c>
<c>NEIGHBOR_TABLE_SIZE</c>
<c>TBD</c>
</texttable>
</section>
<section title="Data Model">
<t>IANA SHALL create/(has created) a "RELOAD Data Model" Registry.
Entries in this registry are 8-bit integers denoting data models, as
described in <xref target="sec-structured-data"></xref>. The
registration policy for this registry is TBD.</t>
<t></t>
<texttable>
<ttcol align="left">Data Model</ttcol>
<ttcol align="right">Identifier</ttcol>
<c>SINGLE_VALUE</c>
<c>TBD</c>
<c>ARRAY</c>
<c>TBD</c>
<c>DICTIONARY</c>
<c>TBD</c>
</texttable>
</section>
<section title="Message Codes">
<t>IANA SHALL create/(has created) a "RELOAD Message Code" Registry.
Entries in this registry are 16-bit integers denoting method codes as
described in <xref target="sec-contents-common-header"></xref> The
registration policy for this registry is TBD.</t>
<t>The initial contents of this registry are:</t>
<t></t>
<texttable>
<ttcol align="left">Message Code Name</ttcol>
<ttcol align="right">Code Value</ttcol>
<c>RESERVED</c>
<c>0</c>
<c>PING_Q</c>
<c>TBD</c>
<c>PING_A</c>
<c>TBD</c>
<c>CONNECT_Q</c>
<c>TBD</c>
<c>CONNECT_A</c>
<c>TBD</c>
<c>TUNNEL_Q</c>
<c>TBD</c>
<c>TUNNEL_A</c>
<c>TBD</c>
<c>STORE_Q</c>
<c>TBD</c>
<c>STORE_A</c>
<c>TBD</c>
<c>FETCH_Q</c>
<c>TBD</c>
<c>FETCH_A</c>
<c>TBD</c>
<c>REMOVE_Q</c>
<c>TBD</c>
<c>REMOVE_A</c>
<c>TBD</c>
<c>FIND_Q</c>
<c>TBD</c>
<c>FIND_A</c>
<c>TBD</c>
<c>JOIN_Q</c>
<c>TBD</c>
<c>JOIN_A</c>
<c>TBD</c>
<c>LEAVE_Q</c>
<c>TBD</c>
<c>LEAVE_A</c>
<c>TBD</c>
<c>UPDATE_Q</c>
<c>TBD</c>
<c>UPDATE_A</c>
<c>TBD</c>
<c>ROUTE_QUERY_Q</c>
<c>TBD</c>
<c>ROUTE_QUERY_A</c>
<c>TBD</c>
<c>RESERVED</c>
<c>0x800..0xffe</c>
<c>ERROR</c>
<c>0xffff</c>
</texttable>
</section>
<t>[[TODO - add IANA registration for p2p_enroll SRV and
p2p_menroll]]</t>
</section>
<section title="Error Codes">
<t>IANA SHALL create/(has created) a "RELOAD Error Code" Registry.
Entries in this registry are 16-bit integers denoting error codes.
[[TODO: Complete this once we decide on error code strategy.</t>
</section>
<section title="Examples">
<t>See draft [TODO add ref] for message flow examples.</t>
</section>
<section title="Acknowledgments">
<t>This draft is a merge of the "REsource LOcation And Discovery
(RELOAD)" draft by David A. Bryan, Marcia Zangrilli and Bruce B.
Lowekamp, the "Address Settlement by Peer to Peer" draft by Cullen
Jennings, Jonathan Rosenberg, and Eric Rescorla, the "Security
Extensions for RELOAD" draft by Bruce B. Lowekamp and James Deverick,
the "A Chord-based DHT for Resource Lookup in P2PSIP" by Marcia
Zangrilli and David A. Bryan, and the Peer-to-Peer Protocol (P2PP) draft
by Salman A. Baset, Henning Schulzrinne, and Marcin Matuszewski.</t>
<t>Thanks to the many people who contributed including: Michael Chen,
TODO - fill in.</t>
</section>
<section title="Appendix: Operation with SIP clients outside the DHT domain"></section>
<section title="Appendix: Notes on DHT Algorithm Selection">
<t>An important point: if you assume NATs are doing ICE to set up
connections, you want a lot fewer connections than you might have on a
very open network - this might push towards something like Chord with
fewer connections than, say, bamboo.</t>
<t>TODO - ref draft-irtf-p2prg-survey-search</t>
</section>
</middle>
<back>
<references title="Normative References">
<reference anchor="RFC2119">
<front>
<title abbrev="RFC Key Words">Key words for use in RFCs to Indicate
Requirement Levels</title>
<author fullname="Scott Bradner" initials="S." surname="Bradner">
<organization>Harvard University</organization>
<address>
<postal>
<street>1350 Mass. Ave.</street>
<street>Cambridge</street>
<street>MA 02138</street>
</postal>
<phone>- +1 617 495 3864</phone>
<email>sob@harvard.edu</email>
</address>
</author>
<date month="March" year="1997" />
<area>General</area>
<keyword>keyword</keyword>
</front>
<seriesInfo name="BCP" value="14" />
<seriesInfo name="RFC" value="2119" />
<format octets="4723" target="ftp://ftp.isi.edu/in-notes/rfc2119.txt"
type="TXT" />
<format octets="15905"
target="http://xml.resource.org/public/rfc/html/rfc2119.html"
type="HTML" />
<format octets="5661"
target="http://xml.resource.org/public/rfc/xml/rfc2119.xml"
type="XML" />
</reference>
<reference anchor="I-D.ietf-mmusic-ice">
<front>
<title>Interactive Connectivity Establishment (ICE): A Protocol for
Network Address Translator (NAT) Traversal for Offer/Answer
Protocols</title>
<author fullname="Jonathan Rosenberg" initials="J"
surname="Rosenberg">
<organization></organization>
</author>
<date day="12" month="June" year="2007" />
<abstract>
<t>This document describes a protocol for Network Address
Translator (NAT) traversal for multimedia sessions established
with the offer/ answer model. This protocol is called Interactive
Connectivity Establishment (ICE). ICE makes use of the Session
Traversal Utilities for NAT (STUN) protocol, applying its binding
discovery and relay usages, in addition to defining a new usage
for checking connectivity between peers. ICE can be used by any
protocol utilizing the offer/answer model, such as the Session
Initiation Protocol (SIP).</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-mmusic-ice-16" />
<format target="http://www.ietf.org/internet-drafts/draft-ietf-mmusic-ice-16.txt"
type="TXT" />
</reference>
<reference anchor="I-D.ietf-behave-rfc3489bis">
<front>
<title>Session Traversal Utilities for (NAT) (STUN)</title>
<author fullname="Jonathan Rosenberg" initials="J"
surname="Rosenberg">
<organization></organization>
</author>
<date day="8" month="March" year="2007" />
<abstract>
<t>Session Traversal Utilities for NAT (STUN) is a lightweight
protocol that serves as a tool for application protocols in
dealing with NAT traversal. It allows a client to determine the IP
address and port allocated to them by a NAT and to keep NAT
bindings open. It can also serve as a check for connectivity
between a client and a server in the presence of NAT, and for the
client to detect failure of the server. STUN works with many
existing NATs, and does not require any special behavior from
them. As a result, it allows a wide variety of applications to
work through existing NAT infrastructure.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft"
value="draft-ietf-behave-rfc3489bis-06" />
<format target="http://www.ietf.org/internet-drafts/draft-ietf-behave-rfc3489bis-06.txt"
type="TXT" />
</reference>
<reference anchor="I-D.ietf-behave-turn">
<front>
<title>Obtaining Relay Addresses from Simple Traversal Underneath
NAT (STUN)</title>
<author fullname="Jonathan Rosenberg" initials="J"
surname="Rosenberg">
<organization></organization>
</author>
<date day="7" month="March" year="2007" />
<abstract>
<t>This specification defines a usage of the Simple Traversal
Underneath NAT (STUN) Protocol for asking the STUN server to relay
packets towards a client. This usage is useful for elements behind
NATs whose mapping behavior is address and port dependent. The
extension purposefully restricts the ways in which the relayed
address can be used. In particular, it prevents users from running
general purpose servers from ports obtained from the STUN
server.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-behave-turn-03" />
<format target="http://www.ietf.org/internet-drafts/draft-ietf-behave-turn-03.txt"
type="TXT" />
</reference>
<reference anchor="I-D.ietf-pkix-cmc-trans">
<front>
<title>Certificate Management over CMS (CMC) Transport
Protocols</title>
<author fullname="Jim Schaad" initials="J" surname="Schaad">
<organization></organization>
</author>
<author fullname="Michael Myers" initials="M" surname="Myers">
<organization></organization>
</author>
<date day="16" month="May" year="2006" />
<abstract>
<t>This document defines a number of transport mechanisms that are
used to move CMC (Certificate Management over CMS (Cryptographic
Message Syntax)) messages. The transport mechanisms described in
this document are: HTTP, file, mail and TCP.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-pkix-cmc-trans-05" />
<format target="http://www.ietf.org/internet-drafts/draft-ietf-pkix-cmc-trans-05.txt"
type="TXT" />
</reference>
<reference anchor="I-D.ietf-pkix-2797-bis">
<front>
<title>Certificate Management Messages over CMS</title>
<author fullname="Michael Myers" initials="M" surname="Myers">
<organization></organization>
</author>
<author fullname="Jim Schaad" initials="J" surname="Schaad">
<organization></organization>
</author>
<date day="3" month="March" year="2006" />
<abstract>
<t>This document defines the base syntax for CMC, a Certificate
Management protocol using CMS (Cryptographic Message Syntax). This
protocol addresses two immediate needs within the Internet PKI
community: 1. The need for an interface to public key
certification products and services based on CMS and PKCS #10
(Public Key Cryptography 2. The need in S/MIME (Secure MIME) for a
certificate enrollment protocol for DSA-signed certificates with
Diffie-Hellman public keys. CMC also requires the use of the
transport document and the requirements usage document along with
this document for a full definition.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-pkix-2797-bis-04" />
<format target="http://www.ietf.org/internet-drafts/draft-ietf-pkix-2797-bis-04.txt"
type="TXT" />
</reference>
<reference anchor="RFC4279">
<front>
<title>Pre-Shared Key Ciphersuites for Transport Layer Security
(TLS)</title>
<author fullname="P. Eronen" initials="P." surname="Eronen">
<organization></organization>
</author>
<author fullname="H. Tschofenig" initials="H." surname="Tschofenig">
<organization></organization>
</author>
<date month="December" year="2005" />
<abstract>
<t>This document specifies three sets of new ciphersuites for the
Transport Layer Security (TLS) protocol to support authentication
based on pre-shared keys (PSKs). These pre-shared keys are
symmetric keys, shared in advance among the communicating parties.
The first set of ciphersuites uses only symmetric key operations
for authentication. The second set uses a Diffie-Hellman exchange
authenticated with a pre-shared key, and the third set combines
public key authentication of the server with pre-shared key
authentication of the client. [STANDARDS TRACK]</t>
</abstract>
</front>
<seriesInfo name="RFC" value="4279" />
<format octets="32160" target="ftp://ftp.isi.edu/in-notes/rfc4279.txt"
type="TXT" />
</reference>
<reference anchor="I-D.ietf-tls-srp">
<front>
<title>Using SRP for TLS Authentication</title>
<author fullname="Dave Taylor" initials="D" surname="Taylor">
<organization></organization>
</author>
<date day="14" month="June" year="2007" />
<abstract>
<t>This memo presents a technique for using the Secure Remote
Password protocol as an authentication method for the Transport
Layer Security protocol.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-tls-srp-14" />
<format target="http://www.ietf.org/internet-drafts/draft-ietf-tls-srp-14.txt"
type="TXT" />
</reference>
<reference anchor="I-D.ietf-mmusic-ice-tcp">
<front>
<title>TCP Candidates with Interactive Connectivity Establishment
(ICE</title>
<author fullname="Jonathan Rosenberg" initials="J"
surname="Rosenberg">
<organization></organization>
</author>
<date day="8" month="March" year="2007" />
<abstract>
<t>Interactive Connectivity Establishment (ICE) defines a
mechanism for NAT traversal for multimedia communication protocols
based on the offer/answer model of session negotiation. ICE works
by providing a set of candidate transport addresses for each media
stream, which are then validated with peer-to-peer connectivity
checks based on Simple Traversal of UDP over NAT (STUN). ICE
provides a general framework for describing alternates, but only
defines UDP-based transport protocols. This specification extends
ICE to TCP-based media, including the ability to offer a mix of
TCP and UDP-based candidates for a single stream.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-mmusic-ice-tcp-03" />
<format target="http://www.ietf.org/internet-drafts/draft-ietf-mmusic-ice-tcp-03.txt"
type="TXT" />
</reference>
<reference anchor="RFC3261">
<front>
<title>SIP: Session Initiation Protocol</title>
<author fullname="J. Rosenberg" initials="J." surname="Rosenberg">
<organization></organization>
</author>
<author fullname="H. Schulzrinne" initials="H."
surname="Schulzrinne">
<organization></organization>
</author>
<author fullname="G. Camarillo" initials="G." surname="Camarillo">
<organization></organization>
</author>
<author fullname="A. Johnston" initials="A." surname="Johnston">
<organization></organization>
</author>
<author fullname="J. Peterson" initials="J." surname="Peterson">
<organization></organization>
</author>
<date month="June" year="2002" />
</front>
<seriesInfo name="RFC" value="3261" />
<format octets="647976"
target="ftp://ftp.isi.edu/in-notes/rfc3261.txt" type="TXT" />
</reference>
<reference anchor="RFC3263">
<front>
<title>Session Initiation Protocol (SIP): Locating SIP
Servers</title>
<author fullname="J. Rosenberg" initials="J." surname="Rosenberg">
<organization></organization>
</author>
<author fullname="H. Schulzrinne" initials="H."
surname="Schulzrinne">
<organization></organization>
</author>
<date month="June" year="2002" />
<abstract>
<t>The Session Initiation Protocol (SIP) uses DNS procedures to
allow a client to resolve a SIP Uniform Resource Identifier (URI)
into the IP address, port, and transport protocol of the next hop
to contact. It also uses DNS to allow a server to send a response
to a backup client if the primary client has failed. This document
describes those DNS procedures in detail. [STANDARDS TRACK]</t>
</abstract>
</front>
<seriesInfo name="RFC" value="3263" />
<format octets="42310" target="ftp://ftp.isi.edu/in-notes/rfc3263.txt"
type="TXT" />
</reference>
<reference anchor="RFC4347">
<front>
<title>Datagram Transport Layer Security</title>
<author fullname="E. Rescorla" initials="E." surname="Rescorla">
<organization></organization>
</author>
<author fullname="N. Modadugu" initials="N." surname="Modadugu">
<organization></organization>
</author>
<date month="April" year="2006" />
<abstract>
<t>This document specifies Version 1.0 of the Datagram Transport
Layer Security (DTLS) protocol. The DTLS protocol provides
communications privacy for datagram protocols. The protocol allows
client/server applications to communicate in a way that is
designed to prevent eavesdropping, tampering, or message forgery.
The DTLS protocol is based on the Transport Layer Security (TLS)
protocol and provides equivalent security guarantees. Datagram
semantics of the underlying transport are preserved by the DTLS
protocol. [STANDARDS TRACK]</t>
</abstract>
</front>
<seriesInfo name="RFC" value="4347" />
<format octets="56014" target="ftp://ftp.isi.edu/in-notes/rfc4347.txt"
type="TXT" />
</reference>
<reference anchor="RFC4828">
<front>
<title>TCP Friendly Rate Control (TFRC): The Small-Packet (SP)
Variant</title>
<author fullname="S. Floyd" initials="S." surname="Floyd">
<organization></organization>
</author>
<author fullname="E. Kohler" initials="E." surname="Kohler">
<organization></organization>
</author>
<date month="April" year="2007" />
<abstract>
<t>This document proposes a mechanism for further experimentation,
but not for widespread deployment at this time in the global
Internet.</t><t> TCP-Friendly Rate Control (TFRC) is a
congestion control mechanism for unicast flows operating in a
best-effort Internet environment (RFC 3448). TFRC was intended for
applications that use a fixed packet size, and was designed to be
reasonably fair when competing for bandwidth with TCP connections
using the same packet size. This document proposes TFRC-SP, a
Small-Packet (SP) variant of TFRC, that is designed for
applications that send small packets. The design goal for TFRC-SP
is to achieve the same bandwidth in bps (bits per second) as a TCP
flow using packets of up to 1500 bytes. TFRC-SP enforces a minimum
interval of 10 ms between data packets to prevent a single flow
from sending small packets arbitrarily
frequently.</t><t> Flows using TFRC-SP compete
reasonably fairly with large-packet TCP and TFRC flows in
environments where large-packet flows and small-packet flows
experience similar packet drop rates. However, in environments
where small-packet flows experience lower packet drop rates than
large-packet flows (e.g., with Drop-Tail queues in units of
bytes), TFRC-SP can receive considerably more than its share of
the bandwidth. This memo defines an Experimental Protocol for the
Internet community.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="4828" />
<format octets="116808"
target="ftp://ftp.isi.edu/in-notes/rfc4828.txt" type="TXT" />
</reference>
</references>
<references title="Informative References">
<reference anchor="I-D.ietf-behave-tcp">
<front>
<title>NAT Behavioral Requirements for TCP</title>
<author fullname="Saikat Guha" initials="S" surname="Guha">
<organization></organization>
</author>
<date day="30" month="April" year="2007" />
<abstract>
<t>This document defines a set of requirements for NATs that
handle TCP that would allow many applications, such as
peer-to-peer applications and on-line games, to work consistently.
Developing NATs that meet this set of requirements will greatly
increase the likelihood that these applications will function
properly.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-behave-tcp-07" />
<format target="http://www.ietf.org/internet-drafts/draft-ietf-behave-tcp-07.txt"
type="TXT" />
</reference>
<reference anchor="I-D.ietf-p2psip-concepts">
<front>
<title>Concepts and Terminology for Peer to Peer SIP</title>
<author fullname="David Bryan" initials="D" surname="Bryan">
<organization></organization>
</author>
<date day="3" month="July" year="2007" />
<abstract>
<t>This document defines concepts and terminology for use of the
Session Initiation Protocol in a peer-to-peer environment where
the traditional proxy-registrar and message routing functions are
replaced by a distributed mechanism that might be implemented
using a distributed hash table or other distributed data mechanism
with similar external properties. This document includes a
high-level view of the functional relationships between the
network elements defined herein, a conceptual model of operations,
and an outline of the related open problems being addressed by the
P2PSIP working group. As this document matures, it is expected to
define the general framework for P2PSIP.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft"
value="draft-ietf-p2psip-concepts-00" />
<format target="http://www.ietf.org/internet-drafts/draft-ietf-p2psip-concepts-00.txt"
type="TXT" />
</reference>
<reference anchor="RFC4145">
<front>
<title>TCP-Based Media Transport in the Session Description Protocol
(SDP)</title>
<author fullname="D. Yon" initials="D." surname="Yon">
<organization></organization>
</author>
<author fullname="G. Camarillo" initials="G." surname="Camarillo">
<organization></organization>
</author>
<date month="September" year="2005" />
<abstract>
<t>This document describes how to express media transport over TCP
using the Session Description Protocol (SDP). It defines the SDP
'TCP' protocol identifier, the SDP 'setup' attribute, which
describes the connection setup procedure, and the SDP 'connection'
attribute, which handles connection reestablishment. [STANDARDS
TRACK]</t>
</abstract>
</front>
<seriesInfo name="RFC" value="4145" />
<format octets="30225" target="ftp://ftp.isi.edu/in-notes/rfc4145.txt"
type="TXT" />
</reference>
<reference anchor="RFC4572">
<front>
<title>Connection-Oriented Media Transport over the Transport Layer
Security (TLS) Protocol in the Session Description Protocol
(SDP)</title>
<author fullname="J. Lennox" initials="J." surname="Lennox">
<organization></organization>
</author>
<date month="July" year="2006" />
<abstract>
<t>This document specifies how to establish secure
connection-oriented media transport sessions over the Transport
Layer Security (TLS) protocol using the Session Description
Protocol (SDP). It defines a new SDP protocol identifier,
'TCP/TLS'. It also defines the syntax and semantics for an SDP
'fingerprint' attribute that identifies the certificate that will
be presented for the TLS session. This mechanism allows media
transport over TLS connections to be established securely, so long
as the integrity of session descriptions is
assured.</t><t> This document extends and updates RFC
4145. [STANDARDS TRACK]</t>
</abstract>
</front>
<seriesInfo name="RFC" value="4572" />
<format octets="38658" target="ftp://ftp.isi.edu/in-notes/rfc4572.txt"
type="TXT" />
</reference>
<reference anchor="RFC2617">
<front>
<title abbrev="HTTP Authentication">HTTP Authentication: Basic and
Digest Access Authentication</title>
<author fullname="John Franks" initials="J." surname="Franks">
<organization>Northwestern University, Department of
Mathematics</organization>
<address>
<postal>
<street>Northwestern University</street>
<city>Evanston</city>
<region>IL</region>
<code>60208-2730</code>
<country>USA</country>
</postal>
<email>john@math.nwu.edu</email>
</address>
</author>
<author fullname="Phillip M. Hallam-Baker" initials="P.M."
surname="Hallam-Baker">
<organization>Verisign Inc.</organization>
<address>
<postal>
<street>301 Edgewater Place</street>
<street>Suite 210</street>
<city>Wakefield</city>
<region>MA</region>
<code>01880</code>
<country>USA</country>
</postal>
<email>pbaker@verisign.com</email>
</address>
</author>
<author fullname="Jeffery L. Hostetler" initials="J.L."
surname="Hostetler">
<organization>AbiSource, Inc.</organization>
<address>
<postal>
<street>6 Dunlap Court</street>
<city>Savoy</city>
<region>IL</region>
<code>61874</code>
<country>USA</country>
</postal>
<email>jeff@AbiSource.com</email>
</address>
</author>
<author fullname="Scott D. Lawrence" initials="S.D."
surname="Lawrence">
<organization>Agranat Systems, Inc.</organization>
<address>
<postal>
<street>5 Clocktower Place</street>
<street>Suite 400</street>
<city>Maynard</city>
<region>MA</region>
<code>01754</code>
<country>USA</country>
</postal>
<email>lawrence@agranat.com</email>
</address>
</author>
<author fullname="Paul J. Leach" initials="P.J." surname="Leach">
<organization>Microsoft Corporation</organization>
<address>
<postal>
<street>1 Microsoft Way</street>
<city>Redmond</city>
<region>WA</region>
<code>98052</code>
<country>USA</country>
</postal>
<email>paulle@microsoft.com</email>
</address>
</author>
<date month="June" year="1999" />
<abstract>
<t>"HTTP/1.0", includes the specification for a Basic Access
Authentication scheme. This scheme is not considered to be a
secure method of user authentication (unless used in conjunction
with some external secure system such as SSL ), as the user name
and password are passed over the network as cleartext.</t>
<t>This document also provides the specification for HTTP's
authentication framework, the original Basic authentication scheme
and a scheme based on cryptographic hashes, referred to as "Digest
Access Authentication". It is therefore also intended to serve as
a replacement for RFC 2069 . Some optional elements specified by
RFC 2069 have been removed from this specification due to problems
found since its publication; other new elements have been added
for compatibility, those new elements have been made optional, but
are strongly recommended.</t>
<t>Like Basic, Digest access authentication verifies that both
parties to a communication know a shared secret (a password);
unlike Basic, this verification can be done without sending the
password in the clear, which is Basic's biggest weakness. As with
most other authentication protocols, the greatest sources of risks
are usually found not in the core protocol itself but in policies
and procedures surrounding its use.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="2617" />
<format octets="77638" target="ftp://ftp.isi.edu/in-notes/rfc2617.txt"
type="TXT" />
<format octets="114842"
target="http://xml.resource.org/public/rfc/html/rfc2617.html"
type="HTML" />
<format octets="87592"
target="http://xml.resource.org/public/rfc/xml/rfc2617.xml"
type="XML" />
</reference>
<reference anchor="RFC2818">
<front>
<title>HTTP Over TLS</title>
<author fullname="E. Rescorla" initials="E." surname="Rescorla">
<organization></organization>
</author>
<date month="May" year="2000" />
<abstract>
<t>This memo describes how to use Transport Layer Security (TLS)
to secure Hypertext Transfer Protocol (HTTP) connections over the
Internet. This memo provides information for the Internet
community.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="2818" />
<format octets="15170" target="ftp://ftp.isi.edu/in-notes/rfc2818.txt"
type="TXT" />
</reference>
<reference anchor="RFC4086">
<front>
<title>Randomness Requirements for Security</title>
<author fullname="D. Eastlake" initials="D." surname="Eastlake">
<organization></organization>
</author>
<author fullname="J. Schiller" initials="J." surname="Schiller">
<organization></organization>
</author>
<author fullname="S. Crocker" initials="S." surname="Crocker">
<organization></organization>
</author>
<date month="June" year="2005" />
<abstract>
<t>Security systems are built on strong cryptographic algorithms
that foil pattern analysis attempts. However, the security of
these systems is dependent on generating secret quantities for
passwords, cryptographic keys, and similar quantities. The use of
pseudo-random processes to generate secret quantities can result
in pseudo-security. A sophisticated attacker may find it easier to
reproduce the environment that produced the secret quantities and
to search the resulting small set of possibilities than to locate
the quantities in the whole of the potential number
space.</t><t> Choosing random quantities to foil a
resourceful and motivated adversary is surprisingly difficult.
This document points out many pitfalls in using poor entropy
sources or traditional pseudo-random number generation techniques
for generating such quantities. It recommends the use of truly
random hardware techniques and shows that the existing hardware on
many systems can be used for this purpose. It provides suggestions
to ameliorate the problem when a hardware solution is not
available, and it gives examples of how large such quantities need
to be for some applications. This document specifies an Internet
Best Current Practices for the Internet Community, and requests
discussion and suggestions for improvements.</t>
</abstract>
</front>
<seriesInfo name="BCP" value="106" />
<seriesInfo name="RFC" value="4086" />
<format octets="114321"
target="ftp://ftp.isi.edu/in-notes/rfc4086.txt" type="TXT" />
</reference>
<reference anchor="RFC3280">
<front>
<title>Internet X.509 Public Key Infrastructure Certificate and
Certificate Revocation List (CRL) Profile</title>
<author fullname="R. Housley" initials="R." surname="Housley">
<organization></organization>
</author>
<author fullname="W. Polk" initials="W." surname="Polk">
<organization></organization>
</author>
<author fullname="W. Ford" initials="W." surname="Ford">
<organization></organization>
</author>
<author fullname="D. Solo" initials="D." surname="Solo">
<organization></organization>
</author>
<date month="April" year="2002" />
<abstract>
<t>This memo profiles the X.509 v3 certificate and X.509 v2
Certificate Revocation List (CRL) for use in the Internet.
[STANDARDS TRACK]</t>
</abstract>
</front>
<seriesInfo name="RFC" value="3280" />
<format octets="295556"
target="ftp://ftp.isi.edu/in-notes/rfc3280.txt" type="TXT" />
</reference>
<!--
<reference anchor="sosimple">
<front>
<title>SOSIMPLE: A Serverless, Standards-based, P2P SIP
Communication System</title>
<author fullname="David A. Bryan" initials="D. A." surname="Bryan">
<organization>College of William and Mary</organization>
</author>
<author fullname="Cullen Jennings" initials="C." surname="Jennings">
<organization>Cisco Systems</organization>
</author>
<author fullname="Bruce B. Lowekamp" initials="B. B."
surname="Lowekamp">
<organization>College of William and Mary</organization>
</author>
<date month="June" year="2005" />
</front>
<seriesInfo name="Proceedings of the 2005 International Workshop on Advanced Architectures and Algorithms for Internet Delivery and Applications (AAA-IDEA)"
value="'05" />
<format target="http://www.cs.wm.edu/~bryan/pubs/bryan-AAA-IDEA2005.pdf"
type="PDF" />
</reference>
<reference anchor="I-D.bryan-p2psip-dsip">
<front>
<title>dSIP: A P2P Approach to SIP Registration and Resource
Location</title>
<author fullname="D. Bryan" initials="D." surname="Bryan">
<organization></organization>
</author>
<author fullname="B. Lowekamp" initials="B." surname="Lowekamp">
<organization></organization>
</author>
<author fullname="C. Jennings" initials="C." surname="Jennings">
<organization></organization>
</author>
<date month="February" year="2007" />
</front>
<seriesInfo name="Internet Draft" value="draft-bryan-p2psip-dsip-00" />
</reference>
-->
<reference anchor="Sybil">
<front>
<title>The Sybil Attack</title>
<author fullname="John R. Douceur" initials="J. R."
surname="Douceur">
<organization>Microsoft Research</organization>
</author>
<date month="March" year="2002" />
</front>
<seriesInfo name="IPTPS" value="02" />
<format target="http://www.cs.rice.edu/Conferences/IPTPS02/101.pdf"
type="PDF" />
</reference>
<reference anchor="Eclipse">
<front>
<title>Eclipse Attacks on Overlay Networks: Threats and
Defenses</title>
<author fullname="Atul Singh" initials="A." surname="Singh">
<organization></organization>
</author>
<author fullname="Tsuen-Wan Ngan" initials="T. W." surname="Ngan">
<organization></organization>
</author>
<author fullname="Peter Druschel" initials="T." surname="Druschel">
<organization></organization>
</author>
<author fullname="Dan S. Wallach" initials="D. S." surname="Wallach">
<organization></organization>
</author>
<date month="April" year="2006" />
</front>
<seriesInfo name="INFOCOM" value="2006" />
</reference>
<reference anchor="I-D.cheshire-dnsext-multicastdns">
<front>
<title>Multicast DNS</title>
<author fullname="Stuart Cheshire" initials="S" surname="Cheshire">
<organization></organization>
</author>
<author fullname="Marc Krochmal" initials="M" surname="Krochmal">
<organization></organization>
</author>
<date day="25" month="August" year="2006" />
</front>
<seriesInfo name="Internet-Draft"
value="draft-cheshire-dnsext-multicastdns-06" />
<format target="http://www.ietf.org/internet-drafts/draft-cheshire-dnsext-multicastdns-06.txt"
type="TXT" />
</reference>
<reference anchor="I-D.cheshire-dnsext-dns-sd">
<front>
<title>DNS-Based Service Discovery</title>
<author fullname="Marc Krochmal" initials="M" surname="Krochmal">
<organization></organization>
</author>
<author fullname="Stuart Cheshire" initials="S" surname="Cheshire">
<organization></organization>
</author>
<date day="28" month="August" year="2006" />
<abstract>
<t>This document describes a convention for naming and structuring
DNS resource records. Given a type of service that a client is
looking for, and a domain in which the client is looking for that
service, this convention allows clients to discover a list of
named instances of that desired service, using only standard DNS
queries. In short, this is referred to as DNS-based Service
Discovery, or DNS-SD.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft"
value="draft-cheshire-dnsext-dns-sd-04" />
<format target="http://www.ietf.org/internet-drafts/draft-cheshire-dnsext-dns-sd-04.txt"
type="TXT" />
</reference>
<!--
<reference anchor='I-D.jennings-p2psip-security'>
<front>
<title>Security Mechanisms for Peer to Peer SIP</title>
<author initials='C' surname='Jennings' fullname='Cullen Jennings'>
<organization />
</author>
<date month='February' day='27' year='2007' />
<abstract><t>This document describes an overview of some security mechanisms for P2P SIP. Specifically it discusses mechanisms that can be used to secure the stored data and the routing in the distributed storage. This draft is an very early draft to outline the possible solution space and far more details would be needed. This work is being discussed on the p2psip@ietf.org mailing list.</t></abstract>
</front>
<seriesInfo name='Internet-Draft' value='draft-jennings-p2psip-security-00' />
<format type='TXT'
target='http://www.ietf.org/internet-drafts/draft-jennings-p2psip-security-00.txt' />
</reference>
-->
<reference anchor="I-D.matthews-p2psip-bootstrap-mechanisms">
<front>
<title>Bootstrap Mechanisms for P2PSIP</title>
<author fullname="Eric Cooper" initials="E" surname="Cooper">
<organization></organization>
</author>
<date day="26" month="February" year="2007" />
<abstract>
<t>This document describes mechanisms that a peer can use to
locate and establish a Peer Protocol connection to an admitting
peer in order to join an overlay network. In the first mechanism,
the joining peer uses multicast to locate a bootstrap peer; in the
second, the node uses one or more bootstrap servers to locate a
bootstrap peer; in both cases, the bootstrap peer then proxies the
request by the joining peer on to the admitting peer. Each
mechanism has its advantages and disadvantages, and a node can
utilize both.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft"
value="draft-matthews-p2psip-bootstrap-mechanisms-00" />
<format target="http://www.ietf.org/internet-drafts/draft-matthews-p2psip-bootstrap-mechanisms-00.txt"
type="TXT" />
</reference>
<reference anchor="I-D.garcia-p2psip-dns-sd-bootstrapping">
<front>
<title>P2PSIP bootstrapping using DNS-SD</title>
<author fullname="Gustavo Garcia" initials="G" surname="Garcia">
<organization></organization>
</author>
<date day="25" month="October" year="2007" />
<abstract>
<t>This document describes a DNS-based bootstrap mechanism to
discover the initial peer or peers needed to join a P2PSIP
Overlay. The document specifies the use of DNS Service Discovery
(DNS-SD) and the format of the required resource records to
support the discovery of P2PSIP peers. This mechanism can be
applied in scenarios with DNS servers or combined with multicast
DNS to fulfill different proposed P2PSIP use cases.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft"
value="draft-garcia-p2psip-dns-sd-bootstrapping-00" />
<format target="http://www.ietf.org/internet-drafts/draft-garcia-p2psip-dns-sd-bootstrapping-00.txt"
type="TXT" />
</reference>
<reference anchor="I-D.camarillo-hip-bone">
<front>
<title>HIP BONE: Host Identity Protocol (HIP) Based Overlay
Networking Environment</title>
<author fullname="Gonzalo Camarillo" initials="G"
surname="Camarillo">
<organization></organization>
</author>
<author fullname="Pekka Nikander" initials="P" surname="Nikander">
<organization></organization>
</author>
<author fullname="Jani Hautakorpi" initials="J" surname="Hautakorpi">
<organization></organization>
</author>
<date day="21" month="December" year="2007" />
<abstract>
<t>This document specifies a framework to build HIP (Host Identity
Protocol)-based overlay networks. This framework uses HIP to
perform connection management. Other functions, such as data
storage and retrieval or overlay maintenance, are implemented
using protocols other than HIP. These protocols are loosely
referred to as peer protocols.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-camarillo-hip-bone-00" />
<format target="http://www.ietf.org/internet-drafts/draft-camarillo-hip-bone-00.txt"
type="TXT" />
</reference>
</references>
</back>
</rfc>| PAFTECH AB 2003-2026 | 2026-04-24 04:07:37 |