RFC: rendezvous point spec

[Roger Dingledine <arma@mit.edu>]

Below is our first draft of a spec for how to rendezvous points with
the current tor architecture. Please comment. :)

                   How to make rendezvous points work

0. Overview

   Rendezvous points are an implementation of location-hidden services
   (server anonymity) in the onion routing network. Location-hidden
   services means Bob can offer a tcp service (say, a webserver) via the
   onion routing network, without revealing the IP of that service.

   The basic idea is to provide censorship resistance for Bob by allowing
   him to advertise a variety of onion routers as his public location
   (nodes known as his Introduction Points, see Section 1). Alice,
   the client, chooses a node known as a Meeting Point (see Section
   2). This extra level of indirection is needed so Bob doesn't serve
   files directly from his public locations (so these nodes don't open
   themselves up to abuse, eg from serving Nazi propaganda in France). The
   extra level of indirection also allows Bob to choose which requests
   to respond to, and which to ignore.

   We provide the necessary glue code so that Alice can view webpages
   on a location-hidden webserver, and Bob can run a location-hidden
   server, with minimal invasive changes (see Section 3). Both Alice
   and Bob must run local onion proxies (OPs) -- software that knows
   how to talk to the onion routing network.

   The big picture follows. We direct the reader to the rest of the
   document for more details and explanation.

   1) Bob chooses some Introduction Points, and advertises them on a
      Distributed Hash Table (DHT).
   2) Bob establishes onion routing connections to each of his
      Introduction Points, and waits.
   3) Alice learns about Bob's service out of band (perhaps Bob gave her
      a pointer, or she found it on a website). She looks up the details
      of Bob's service from the DHT.
   4) Alice chooses and establishes a Meeting Point for this transaction.
   5) Alice goes to one of Bob's Introduction Points, and gives it a blob
      (encrypted for Bob) which tells him about herself and the Meeting
      Point she chose. The Introduction Point sends the blob to Bob.
   6) Bob chooses whether to ignore the blob, or to onion route to MP.
      Let's assume the latter.
   7) MP plugs together Alice and Bob. Note that MP doesn't know (or care)
      who Alice is, or who Bob is; and it can't read anything they
      transmit either, because they share a session key.
   8) Alice sends a 'begin' cell along the circuit. It makes its way
      to Bob's onion proxy. Bob's onion proxy connects to Bob's webserver.
   9) Data goes back and forth as usual.

1. Introduction service

   Bob wants to learn about client requests for communication, but
   wants to avoid responding unnecessarily to unauthorized clients.
   Bob's proxy opens a circuit, and tells some onion router on that
   circuit to expect incoming connections, and notify Bob of them.

   When establishing such an introduction point, Bob provides the onion
   router with a public "introduction" key.  The hash of this public
   key identifies a unique Bob, and (since Bob is required to sign his
   messages) prevents anybody else from usurping Bob's introduction
   point in the future.  Additionally, Bob can use the same public key
   to establish an introduction point on another onion router (OR),
   and Alice can still be confident that Bob is the same server.

   (The set-up-an-introduction-point command should come via a
   RELAY_BIND_INTRODUCTION cell. This cell creates a new stream on the
   circuit from Bob to the introduction point.)

   ORs that support introduction run an introduction service on a
   separate port.  When Alice wants to notify Bob of a meeting point,
   she connects (directly or via Tor) to the introduction port, and
   sends the following:

     MEETING REQUEST
        RSA-OAEP encrypted with server's public key:
[20 bytes] Hash of Bob's public key (identifies which Bob to notify)
[ 0 bytes] Initial authentication [optional]
        RSA encrypted with Bob's public key:
[16 bytes] Symmetric key for encrypting blob past RSA
[ 6 bytes] Meeting point (IP/port)
[ 8 bytes] Meeting cookie
[ 0 bytes] End-to-end authentication [optional]
[98 bytes] g^x part 1 (inside the RSA)
[30 bytes] g^x part 2 (symmetrically encrypted)

   The meeting point and meeting cookie allow Bob to contact Alice and
   prove his identity; the end-to-end authentication enables Bob to
   decide whether to talk to Alice; the initial authentication enables
   the meeting point to pre-screen introduction requests before sending
   them to Bob.  (See Section 2 for a discussion of meeting points;
   see Section 1.1 for an example authentication mechanism.)

   The authentication steps are the appropriate places for the
   introduction server or Bob to do replay prevention, if desired.

   When the introduction point receives a valid meeting request, it
   sends the portion intended for Bob along the stream
   created by Bob's RELAY_BIND_INTRODUCTION.  Bob then, at his
   discretion, connects to Alice's meeting point.

1.1. An example authentication scheme for introduction services

   Bob makes two short-term secrets SB and SN, and tells the
   introduction point about SN.  Bob gives Alice a cookie consisting
   of A,B,C such that H(A|SB)=B and H(A|SN)=C.  Alice's initial
   authentication is <A,C>; Alice's end-to-end authentication is <A,B>.

   [Maybe] Bob keeps a replay cache of A values, and doesn't allow any
   value to be used twice.  Over time, Bob rotates SB and SN.

   [Maybe] Each 'A' has an expiration time built in to it.

   In reality, we'll want to pick a scheme that (a) wasn't invented from
   scratch in an evening, and (b) doesn't require Alice to remember this
   many bits (see section 3.2).

2. Meeting points

   For Bob to actually reply to Alice, Alice first establishes a
   circuit to an onion router R, and sends a RELAY_BIND_MEETING cell
   to that onion router.  The RELAY_BIND_MEETING cell contains a
   'Meeting cookie' (MC) that Bob can use to authenticate to R.  R
   remembers the cookie and associates it with Alice.

   Later, Bob also routes to R and sends R a RELAY_JOIN_MEETING cell with
   the meeting cookie MC.  After this point, R routes all traffic from
   Bob's circuit or Alice's circuit as if the two circuits were joined:
   any RELAY cells that are not for a recognized topic are passed down
   Alice or Bob's circuit.  Bob's first cell to Alice contains g^y.

   To prevent R from reading their traffic, Alice and Bob derive two
   end-to-end keys from g^{xy}, and they each treat R as just another
   hop on the circuit.  (These keys are used in addition to the series
   of encryption keys already in use on Alice and Bob's circuits.)

   Bob's OP accepts RELAY_BEGIN, RELAY_DATA, RELAY_END, and
   RELAY_SENDME cells from Alice.  Alice's OP accepts RELAY_DATA,
   RELAY_END, and RELAY_SENDME cells from Bob.  All RELAY_BEGIN cells
   to Bob must have target IP and port of zero; Bob's OP will redirect
   them to the actual target IP and port of Bob's server.

   Alice and Bob's OPs disallow CREATE or RELAY_EXTEND cells as usual.

3. Application interface

3.1. Application interface: server side

   Bob has a service that he wants to offer to the world but keep its
   location hidden.  He configures his local OP to know about this
   service, including the following data:

     Local IP and port of the service
     Strategy for choosing introduction points
       (for now, just randomly pick among the ORs offering it)
     Strategy for user authentication
       (for now, just accept all users)
     Public (RSA) key (one for each service Bob offers)

   Bob chooses a set of N Introduction servers on which to advertise
   his service.

   We assume the existence of a robust decentralized efficient lookup
   system (call it "DHT" for distributed hash table -- note that the
   onion routers can run nodes). Bob publishes
     * Bob's Public Key for that service
     * Expiration date ("don't use after")
     * Introduction server 0 ... Introduction server N
     (All signed by Bob's Public Key)
   into DHT, indexed by the hash of Bob's Public Key. Bob should
   periodically republish his introduction information with a new
   expiration date (and possibly with new/different introduction servers
   if he wants), so Alice can trust that DHT is giving her an up-to-date
   version. The Chord CFS paper describes a sample DHT that allows
   authenticated updating.

3.2. Application interface: client side

   We require that the client interface remain a SOCKS proxy, and we
   require that Alice shouldn't have to modify her applications. Thus
   we encode all of the necessary information into the hostname (more
   correctly, fqdn) that Alice uses, eg when clicking on a url in her
   browser.

   To establish a connection to Bob, Alice needs to know an Introduction
   point, Bob's PK, and some authentication cookie. Because encoding this
   information into the hostname will be too long for a typical hostname,
   we instead use a layer of indirection. We encode a hash of Bob's PK
   (10 bytes is sufficient since we're not worrying about collisions),
   and also the authentication token (empty for now). Location-hidden
   services use the special top level domain called '.onion': thus
   hostnames take the form x.y.onion where x is the hash of PK, and y
   is the authentication cookie. If no cookie is required, the hostname
   can simply be of the form x.onion. Assuming only case insensitive
   alphanumeric and hyphen, we get a bit more than 6 bits encoded
   per character, meaning the x part of the hostname will be about
   13 characters.

   Alice's onion proxy examines hostnames and recognizes when they're
   destined for a hidden server. If so, it decodes the PK and performs
   the steps in Section 0 above.