~bzr-pqm/bzr/bzr.dev

================
Network Protocol
================

:Date: 2009-01-07


.. contents::


Overview
========

The smart protocol provides a way to send a requests and corresponding
responses to communicate with a remote bzr process.

Layering
========

Medium
------

At the bottom level there is either a socket, pipes, or an HTTP
request/response.  We call this layer the *medium*.  It is responsible for
carrying bytes between a client and server.  For sockets, we have the idea
that you have multiple requests and get a read error because the other
side did shutdown.  For pipes we have read pipe which will have a zero
read which marks end-of-file.  For HTTP server environment there is no
end-of-stream because each request coming into the server is independent.

So we need a wrapper around pipes and sockets to separate out requests
from substrate and this will give us a single model which is consistent
for HTTP, sockets and pipes.

Protocol
--------

On top of the medium is the *protocol*.  This is the layer that
deserialises bytes into the structured data that requests and responses
consist of.

Request/Response processing
---------------------------

On top of the protocol is the logic for processing requests (on the
server) or responses (on the client).

Server-side
-----------

Sketch::

 MEDIUM  (factory for protocol, reads bytes & pushes to protocol,
          uses protocol to detect end-of-request, sends written
          bytes to client) e.g. socket, pipe, HTTP request handler.
  ^
  | bytes.
  v

 PROTOCOL(serialization, deserialization)  accepts bytes for one
          request, decodes according to internal state, pushes
          structured data to handler.  accepts structured data from
          handler and encodes and writes to the medium.  factory for
          handler.
  ^
  | structured data
  v

 HANDLER  (domain logic) accepts structured data, operates state
          machine until the request can be satisfied,
          sends structured data to the protocol.

Request handlers are registered in the `bzrlib.smart.request` module.


Client-side
-----------

Sketch::

 CLIENT   domain logic, accepts domain requests, generated structured
          data, reads structured data from responses and turns into
          domain data.  Sends structured data to the protocol.
          Operates state machines until the request can be delivered
          (e.g. reading from a bundle generated in bzrlib to deliver a
          complete request).

          This is RemoteBzrDir, RemoteRepository, etc.
  ^
  | structured data
  v

 PROTOCOL  (serialization, deserialization)  accepts structured data for one
          request, encodes and writes to the medium.  Reads bytes from the
          medium, decodes and allows the client to read structured data.
  ^
  | bytes.
  v

 MEDIUM   accepts bytes from the protocol & delivers to the remote server.
          Allows the protocol to read bytes e.g. socket, pipe, HTTP request.

The domain logic is in `bzrlib.remote`: `RemoteBzrDir`, `RemoteBranch`,
and so on.

There is also an plain file-level transport that calls remote methods to
manipulate files on the server in `bzrlib.transport.remote`.

Protocol description
====================

Version one
-----------

Version one of the protocol was introduced in Bazaar 0.11.

The protocol (for both requests and responses) is described by::

  REQUEST := MESSAGE_V1
  RESPONSE := MESSAGE_V1
  MESSAGE_V1 := ARGS [BODY]

  ARGS := ARG [MORE_ARGS] NEWLINE
  MORE_ARGS := SEP ARG [MORE_ARGS]
  SEP := 0x01

  BODY := LENGTH NEWLINE BODY_BYTES TRAILER
  LENGTH := decimal integer
  TRAILER := "done" NEWLINE

That is, a tuple of arguments separated by Ctrl-A and terminated with a
newline, followed by length prefixed body with a constant trailer.  Note
that although arguments are not 8-bit safe (they cannot include 0x01 or
0x0a bytes without breaking the protocol encoding), the body is.

Version two
-----------

Version two was introduced in Bazaar 0.16.

The request protocol is::

  REQUEST_V2 := "bzr request 2" NEWLINE MESSAGE_V2

The response protocol is::

  RESPONSE_V2 := "bzr response 2" NEWLINE RESPONSE_STATUS NEWLINE MESSAGE_V2
  RESPONSE_STATUS := "success" | "failed"

Future versions should follow this structure, like version two does::

  FUTURE_MESSAGE := VERSION_STRING NEWLINE REST_OF_MESSAGE

This is so that clients and servers can read bytes up to the first newline
byte to determine what version a message is.

For compatibility will all versions (past and future) of bzr clients,
servers that receive a request in an unknown protocol version should
respond with a single-line error terminated with 0x0a (NEWLINE), rather
than structured response prefixed with a version string.

Version two of the message protocol is::

  MESSAGE_V2 := ARGS [BODY_V2]
  BODY_V2 := BODY | STREAMED_BODY

That is, a version one length-prefixed body, or a version two streamed
body.

Version two with streamed bodies
--------------------------------

An extension to version two allows streamed bodies.  A streamed body looks
a lot like HTTP's chunked encoding::
 
  STREAMED_BODY := "chunked" NEWLINE CHUNKS TERMINATOR
  CHUNKS := CHUNK [CHUNKS]
  CHUNK := HEX_LENGTH CHUNK_CONTENT
  HEX_LENGTH := HEX_DIGITS NEWLINE
  CHUNK_CONTENT := bytes
  
  TERMINATOR := SUCCESS_TERMINATOR | ERROR_TERMINATOR
  SUCCESS_TERMINATOR := 'END' NEWLINE
  ERROR_TERMINATOR := 'ERR' NEWLINE CHUNKS SUCCESS_TERMINATOR

That is, the body consists of a series of chunks.  Each chunk starts with
a length prefix in hexadecimal digits, followed by an ASCII newline byte.
The end of the body is signaled by '``END\\n``', or by '``ERR\\n``'
followed by error args, one per chunk.  Note that these args are 8-bit
safe, unlike request args.

A streamed body starts with the string "chunked" so that legacy clients
and servers will not mistake the first chunk as the start of a version one
body.

The type of body (length-prefixed or chunked) in a response is always the
same for a given request method.  Only new request methods introduced in
Bazaar 0.91 and later use streamed bodies.

Version three
-------------

.. note::
  
  For some discussion of the requirements that led to this new protocol
  version, see `bug #83935`_.

.. _bug #83935: https://bugs.launchpad.net/bzr/+bug/83935

Version three has bencoding of most protocol structures, to make parsing
simpler.  For extra parsing convenience, these structures are length
prefixed::

  LENGTH_PREFIX := 32-bit unsigned integer in network byte order

Unlike earlier versions, clients and servers are no longer required to
know which request verbs and responses will have bodies attached.  Because
of length-prefixing and other changes, it is always possible to know when
a complete request or response has been read, even if the server
implements no verbs.

The underlying message format is::

  MESSAGE := MAGIC NEWLINE HEADERS CONTENTS END_MESSAGE
  MAGIC := "bzr message 3 (bzr 1.6)"
  HEADERS := LENGTH_PREFIX bencoded_dict
  END_MESSAGE := "e"

  BODY := MESSAGE_PART+ 
  MESSAGE_PART := ONE_BYTE | STRUCTURE | BYTES
  ONE_BYTE := "o" byte
  STRUCTURE := "s" LENGTH_PREFIX bencoded_structure
  BYTES := "b" LENGTH_PREFIX bytes

(Where ``+`` indicates one or more.)

This format allows an arbitrary sequence of message parts to be encoded
in a single message.  The contents of a MESSAGE have a higher-level
message, but knowing just this amount of data it's possible to
deserialize and consume a message, so that implementations can respond to
messages sent by later versions.

Headers
~~~~~~~

Each request and response will have “headers”, a dictionary of key-value pairs.
The keys must be strings, not any other type of value.

Currently, the only defined header is “Software version”.  Both the client and
the server should include a “Software version” header, with a value of a
free-form string such as “bzrlib 1.5”, to aid debugging and logging.  Clients
and servers **should not** vary behaviour based on this string.

Conventional requests and responses
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

By convention, most requests and responses have a simple “arguments plus
optional body” structure, as in earlier protocol versions.  This section
describes how such messages are encoded.  All requests and responses
defined by earlier protocol versions must be encoded in this way.

Conventional requests will send a CONTENTS of ::

  CONV_REQ := ARGS SINGLE_OR_STREAMED_BODY?
  SINGLE_OR_STREAMED_BODY := BYTES 
        | BYTES+ TRAILER
         
  ARGS := STRUCTURE(argument_tuple) 
  TRAILER := SUCCESS_STATUS | ERROR
  SUCCESS_STATUS := ONE_BYTE("S")
  ERROR := ONE_BYTE("E") STRUCTURE(argument_tuple)

Conventional responses will send CONTENTS of ::

  CONV_RESP := RESP_STATUS ARGS SINGLE_OR_STREAMED_BODY?
  RESP_STATUS := ONE_BYTE("S") | ONE_BYTE("E")

If the RESP_STATUS is success ("S"), the arguments are the
method-dependent result.  

For errors (where the Status byte of a response or a streamed body is
"E"), the situation is analagous to requests.  The first item in the
encoded sequence must be a string of the error name.  The other arguments
supply details about the error, and their number and types will depend on
the type of error (as identified by the error name).

Note that the streamed body from version two is now just multiple
BYTES parts.

The end of the request or response is indicated by the lower-level 
END_MESSAGE.  If there's only one BYTES element in the body, the TRAILER
may or may not be present, depending on whether it was sent as a single
chunk or as a stream that happens to have one element.

  *(Discussion)* The success marker at the end of a streamed body seems
  redundant; it doesn't have space for any arguments, and the end of the
  body is marked anyhow by the end of the message.  Recipients shouldn't
  take any action on it, though they should map an error into raising an
  error locally.

  1.10 clients don't assert that they get a status byte at the end of the
  message.  They will complain (in
  ``ConventionalResponseHandler.byte_part_received``) if they get an
  initial success and then another byte part with no intervening bytes.
  If we stop sending the final success message and only flag errors
  they'll only get one if the error is detected after streaming starts but
  before any bytes are actually sent.  Possibly we should wait until at 
  least the first chunk is ready before declaring success.

For new methods, these sequences are just a convention and may be varied
if appropriate for a particular request or response.  However, each
request should at least start with a STRUCTURE encoding the arguments
tuple.  The first element of that tuple must be a string that names the
request method.  (Note that arguments in this protocol version are
bencoded.  As a result, unlike previous protocol versions, arguments in
this version are 8-bit clean.)

  (Discussion) We're discussing having the byte segments be not just a
  method for sending a stream across the network, but actually having them
  be preserved in the rpc from end to end.  This may be useful when
  there's an iterator on one side feeding in to an iterator on the other,
  if it avoids doing chunking and byte-counting at two levels, and if
  those iterators are a natural place to get good granularity.  Also, for 
  cases like ``insert_record_stream`` the server can't do much with the
  data until it gets a whole chunk, and so it'll be natural and efficient
  for it to be called with one chunk at a time.

  On the other hand, there may be times when we've got some bytes from the 
  network but not a full chunk, and it might be worthwhile to pass it up.
  If we promise to preserve chunks, then to do this we'd need two separate
  streaming interfaces: "we got a chunk" and "we got some bytes but not
  yet a full chunk".  For ``insert_record_stream`` the second might not be
  useful, but it might be good when writing to a file where any number of
  bytes can be processed.

  If we promise to preserve chunks, it'll tend to make some RPCs work only
  in chunks, and others just on whole blocks, and we can't so easily
  migrate RPCs from one to the other transparently to older
  implementations.

  The data inside those chunks will be serialized anyhow, and possibly the
  data inside them will already be able to be serialized apart without
  understanding the chunks.  Also, we might want to use these formats e.g.
  for pack files or in bundles, and so they don't particularly need
  lower-level chunking.  So the current (unmerged, unstable) record stream
  serialization turns each record into a bencoded tuple and it'd be
  feasible to parse one tuple at a time from a byte stream that contains a
  sequence of them.

  So we've decided that the chunks won't be semantic, and code should not
  count on them being preserved from client to server.

Early error returns
~~~~~~~~~~~~~~~~~~~

  *(Discussion)* It would be nice if the server could notify the client of
  errors even before a streaming request has finished.  This could cover
  situtaions such as the server not understanding the request, it being
  unable to open the requested location, or it finding that some of the
  revisions being sent are not actually needed.

  Especially in the last case, we'd like to be able to gracefully notice
  the condition while the client is writing, and then have it adapt its
  behaviour.  In any case, we don't want to have drop and restart the
  network stream.

  It should be possible for the client to finish its current chunk and
  then its message, possibly with an error to cancel what's already been
  sent.

  This relies on the client being able to read back from the server while
  it's writing.  This is technically difficult for http but feasible over
  a socket or ssh.

  We'd need a clean way to pass this back to the request method, even
  though it's presumably in the middle of doing its body iterator.
  Possibly the body iterator could be manually given a reference to the
  request object, and it can poll it to see if there's a response.

  Perhaps we need to distinguish error conditions, which should turn into
  a client-side error regardless of the request code, from early success,
  which should be handled only if the request code specifically wants to
  do it.

Full-duplex operation
~~~~~~~~~~~~~~~~~~~~~

  Code not geared to do pipelined requests, and this might require doing
  asynchrony within bzrlib.  We might want to either go fully pipelined
  and asynchronous, but there might be a profitable middle ground.

  The particular case where duplex communication would be good is in
  working towards the common points in the graphs between the client and
  server: we want to send speculatively, but detect as soon as they've
  matched up.

  So we could for instance have a synchronous core, but rely on the OS
  network buffering to allow us to work on batches of say 64kB.  We can
  also pipeline requests and responses, without allowing for them
  happening out of order, or mixed requests happening at the same time.

  Wonder how our network performance would have turned out now if we'd
  done full-duplex from the start, and ignored hpss over http.  We have
  pretty good (readonly) http support just over dumb http, and that may be
  better for many users.



APIs
====

On the client, the bzrlib code is "in charge": when it makes a request, or
asks from data from the network, that causes network IO.  The server is
event driven: the network code tells the response handler when data has
been received, and it takes back a Response object from the request
handler that is then polled for body stream data.

Paths
=====

Paths are passed across the network.  The client needs to see a namespace
that includes any repository that might need to be referenced, and the
client needs to know about a root directory beyond which it cannot ascend.

Servers run over ssh will typically want to be able to access any path the
user can access.  Public servers on the other hand (which might be over
http, ssh or tcp) will typically want to restrict access to only a
particular directory and its children, so will want to do a software
virtual root at that level.  In other words they'll want to rewrite
incoming paths to be under that level (and prevent escaping using ../
tricks).  The default implementation in bzrlib does this using the
`bzrlib.transport.chroot` module.

URLs that include ~ should probably be passed across to the server
verbatim and the server can expand them.  This will proably not be
meaningful when limited to a directory?  See `bug 109143`_.

.. _bug 109143: https://bugs.launchpad.net/bzr/+bug/109143


Requests
========

The first argument of a request specifies the request method.

The available request methods are registered in `bzrlib.smart.request`.

**XXX**: ideally the request methods should be documented here.
Contributions welcome!


Recognised errors
=================

The first argument of an error response specifies the error type.

One possible error name is ``UnknownMethod``, which means the server does
not recognise the verb used by the client's request.  This error was
introduced in version three.

**XXX**: ideally the error types should be documented here.  Contributions
welcome!

..
   vim: ft=rst tw=74 ai