<?xml version="1.0"?>

<!DOCTYPE modulesynopsis SYSTEM "/style/modulesynopsis.dtd">

<?xml-stylesheet type="text/xsl" href="/style/manual.en.xsl"?>

<modulesynopsis metafile="mod_proxy_ajp.xml.meta">

<name>mod_proxy_ajp</name>

<description>AJP support module for

<module>mod_proxy</module></description>

<status>Extension</status>

<sourcefile>proxy_ajp.c</sourcefile>

<identifier>proxy_ajp_module</identifier>

<compatibility>Available in version 2.1 and later</compatibility>

<summary>

<p>This module <em>requires</em> the service of <module

>mod_proxy</module>. It provides support for the

<code>Apache JServ Protocol version 1.3</code> (hereafter

<em>AJP13</em>).</p>

<p>Thus, in order to get the ability of handling <code>AJP13</code>

protocol, <module>mod_proxy</module> and

<module>mod_proxy_ajp</module> have to be present in the server.</p>

<note type="warning"><title>Warning</title>

<p>Do not enable proxying until you have <a

href="mod_proxy.html#access">secured your server</a>. Open proxy

servers are dangerous both to your network and to the Internet at

large.</p>

</note>

</summary>

<seealso><module>mod_proxy</module></seealso>

<section id="overviewprotocol"><title>Overview of the protocol</title>

<p>The <code>AJP13</code> protocol is packet-oriented. A binary format

was presumably chosen over the more readable plain text for reasons of

performance. The web server communicates with the servlet container over

TCP connections. To cut down on the expensive process of socket creation,

the web server will attempt to maintain persistent TCP connections to the

servlet container, and to reuse a connection for multiple request/response

cycles.</p>

<p>Once a connection is assigned to a particular request, it will not be

used for any others until the request-handling cycle has terminated. In

other words, requests are not multiplexed over connections. This makes

for much simpler code at either end of the connection, although it does

cause more connections to be open at once.</p>

<p>Once the web server has opened a connection to the servlet container,

the connection can be in one of the following states:</p>

<ul>

<li> Idle <br/> No request is being handled over this connection. </li>

<li> Assigned <br/> The connecton is handling a specific request.</li>

</ul>

<p>Once a connection is assigned to handle a particular request, the basic

request informaton (e.g. HTTP headers, etc) is sent over the connection in

a highly condensed form (e.g. common strings are encoded as integers).

Details of that format are below in Request Packet Structure. If there is a

body to the request <code>(content-length > 0)</code>, that is sent in a

separate packet immediately after.</p>

<p>At this point, the servlet container is presumably ready to start

processing the request. As it does so, it can send the

following messages back to the web server:</p>

<ul>

<li>SEND_HEADERS <br/>Send a set of headers back to the browser.</li>

<li>SEND_BODY_CHUNK <br/>Send a chunk of body data back to the browser.

</li>

<li>GET_BODY_CHUNK <br/>Get further data from the request if it hasn't all

been transferred yet. This is necessary because the packets have a fixed

maximum size and arbitrary amounts of data can be included the body of a

request (for uploaded files, for example). (Note: this is unrelated to

HTTP chunked tranfer).</li>

<li>END_RESPONSE <br/> Finish the request-handling cycle.</li>

</ul>

<p>Each message is accompanied by a differently formatted packet of data.

See Response Packet Structures below for details.</p>

</section>

<section id="basppacketstruct"><title>Basic Packet Structure</title>

<p>There is a bit of an XDR heritage to this protocol, but it differs

in lots of ways (no 4 byte alignment, for example).</p>

<p>Byte order: I am not clear about the endian-ness of the individual

bytes. I'm guessing the bytes are little-endian, because that's what

XDR specifies, and I'm guessing that sys/socket library is magically

making that so (on the C side). If anyone with a better knowledge of

socket calls can step in, that would be great.</p>

<p>There are four data types in the protocol: bytes, booleans,

integers and strings.</p>

<dl>

<dt><strong>Byte</strong></dt><dd>A single byte.</dd>

<dt><strong>Boolean</strong></dt>

<dd>A single byte, <code>1 = true</code>, <code>0 = false</code>.

Using other non-zero values as true (i.e. C-style) may work in some places,

but it won't in others.</dd>

<dt><strong>Integer</strong></dt>

<dd>A number in the range of <code>0 to 2^16 (32768)</code>. Stored in

2 bytes with the high-order byte first.</dd>

<dt><strong>String</strong></dt>

<dd>A variable-sized string (length bounded by 2^16). Encoded with

the length packed into two bytes first, followed by the string

(including the terminating '\0'). Note that the encoded length does

<strong>not</strong> include the trailing '\0' -- it is like

<code>strlen</code>. This is a touch confusing on the Java side, which

is littered with odd autoincrement statements to skip over these

terminators. I believe the reason this was done was to allow the C

code to be extra efficient when reading strings which the servlet

container is sending back -- with the terminating \0 character, the

C code can pass around references into a single buffer, without copying.

if the \0 was missing, the C code would have to copy things out in order

to get its notion of a string.</dd>

</dl>

<section><title>Packet Size</title>

<p>According to much of the code, the max packet size is <code>

8 * 1024 bytes (8K)</code>. The actual length of the packet is encoded in

the header.</p>

</section>

<section><title>Packet Headers</title>

<p>Packets sent from the server to the container begin with

<code>0x1234</code>. Packets sent from the container to the server

begin with <code>AB</code> (that's the ASCII code for A followed by the

ASCII code for B). After those first two bytes, there is an integer

(encoded as above) with the length of the payload. Although this might

suggest that the maximum payload could be as large as 2^16, in fact, the

code sets the maximum to be 8K.</p>

<table>

<tr>

<td colspan="6"><em>Packet Format (Server->Container)</em></td>

</tr>

<tr>

<td>Byte</td>

<td>0</td>

<td>1</td>

<td>2</td>

<td>3</td>

<td>4...(n+3)</td>

</tr>

<tr>

<td>Contents</td>

<td>0x12</td>

<td>0x34</td>

<td colspan="2">Data Length (n)</td>

<td>Data</td>

</tr>

</table>

<table>

<tr>

<td colspan="6"><em>Packet Format (Container->Server)</em></td>

</tr>

<tr>

<td>Byte</td>

<td>0</td>

<td>1</td>

<td>2</td>

<td>3</td>

<td>4...(n+3)</td>

</tr>

<tr>

<td>Contents</td>

<td>A</td>

<td>B</td>

<td colspan="2">Data Length (n)</td>

<td>Data</td>

</tr>

</table>

<p>For most packets, the first byte of the payload encodes the type of

message. The exception is for request body packets sent from the server to

the container -- they are sent with a standard packet header (<code>

0x1234</code> and then length of the packet), but without any prefix code

after that.</p>

<p>The web server can send the following messages to the servlet

container:</p>

<table>

<tr>

<td>Code</td>

<td>Type of Packet</td>

<td>Meaning</td>

</tr>

<tr>

<td>2</td>

<td>Forward Request</td>

<td>Begin the request-processing cycle with the following data</td>

</tr>

<tr>

<td>7</td>

<td>Shutdown</td>

<td>The web server asks the container to shut itself down.</td>

</tr>

<tr>

<td>8</td>

<td>Ping</td>

<td>The web server asks the container to take control

(secure login phase).</td>

</tr>

<tr>

<td>10</td>

<td>CPing</td>

<td>The web server asks the container to respond quickly with a CPong.

</td>

</tr>

<tr>

<td>none</td>

<td>Data</td>

<td>Size (2 bytes) and corresponding body data.</td>

</tr>

</table>

<p>To ensure some basic security, the container will only actually do the

<code>Shutdown</code> if the request comes from the same machine on which

it's hosted.</p>

<p>The first <code>Data</code> packet is send immediatly after the

<code>Forward Request</code> by the web server.</p>

<p>The servlet container can send the following types of messages to the

webserver:</p>

<table>

<tr>

<td>Code</td>

<td>Type of Packet</td>

<td>Meaning</td>

</tr>

<tr>

<td>3</td>

<td>Send Body Chunk</td>

<td>Send a chunk of the body from the servlet container to the web

server (and presumably, onto the browser). </td>

</tr>

<tr>

<td>4</td>

<td>Send Headers</td>

<td>Send the response headers from the servlet container to the web

server (and presumably, onto the browser).</td>

</tr>

<tr>

<td>5</td>

<td>End Response</td>

<td>Marks the end of the response (and thus the request-handling cycle).

</td>

</tr>

<tr>

<td>6</td>

<td>Get Body Chunk</td>

<td>Get further data from the request if it hasn't all been

transferred yet.</td>

</tr>

<tr>

<td>9</td>

<td>CPong Reply</td>

<td>The reply to a CPing request</td>

</tr>

</table>

<p>Each of the above messages has a different internal structure, detailed

below.</p>

</section>

</section>

<section id="rpacetstruct"><title>Request Packet Structure</title>

<p>For messages from the server to the container of type

<em>Forward Request</em>:</p>

<example><pre>

AJP13_FORWARD_REQUEST :=

prefix_code (byte) 0x02 = JK_AJP13_FORWARD_REQUEST

method (byte)

protocol (string)

req_uri (string)

remote_addr (string)

remote_host (string)

server_name (string)

server_port (integer)

is_ssl (boolean)

num_headers (integer)

request_headers *(req_header_name req_header_value)

attributes *(attribut_name attribute_value)

request_terminator (byte) OxFF

</pre></example>

<p>The <code>request_headers</code> have the following structure:

</p><example><pre>

req_header_name :=

sc_req_header_name | (string) [see below for how this is parsed]

sc_req_header_name := 0xA0xx (integer)

req_header_value := (string)

</pre></example>

<p>The <code>attributes</code> are optional and have the following

structure:</p>

<example><pre>

attribute_name := sc_a_name | (sc_a_req_attribute string)

attribute_value := (string)

</pre></example>

<p>Not that the all-important header is <code>content-length</code>,

because it determines whether or not the container looks for another

packet immediately.</p>

<section><title>Detailed description of the elements of Forward Request

</title></section>

<section><title>Request prefix</title>

<p>For all requests, this will be 2. See above for details on other Prefix

codes.</p>

</section>

<section><title>Method</title>

<p>The HTTP method, encoded as a single byte:</p>

<table>

<tr><td>Command Name</td><td>Code</td></tr>

<tr><td>OPTIONS</td><td>1</td></tr>

<tr><td>GET</td><td>2</td></tr>

<tr><td>HEAD</td><td>3</td></tr>

<tr><td>POST</td><td>4</td></tr>

<tr><td>PUT</td><td>5</td></tr>

<tr><td>DELETE</td><td>6</td></tr>

<tr><td>TRACE</td><td>7</td></tr>

<tr><td>PROPFIND</td><td>8</td></tr>

<tr><td>PROPPATCH</td><td>9</td></tr>

<tr><td>MKCOL</td><td>10</td></tr>

<tr><td>COPY</td><td>11</td></tr>

<tr><td>MOVE</td><td>12</td></tr>

<tr><td>LOCK</td><td>13</td></tr>

<tr><td>UNLOCK</td><td>14</td></tr>

<tr><td>ACL</td><td>15</td></tr>

<tr><td>REPORT</td><td>16</td></tr>

<tr><td>VERSION-CONTROL</td><td>17</td></tr>

<tr><td>CHECKIN</td><td>18</td></tr>

<tr><td>CHECKOUT</td><td>19</td></tr>

<tr><td>UNCHECKOUT</td><td>20</td></tr>

<tr><td>SEARCH</td><td>21</td></tr>

<tr><td>MKWORKSPACE</td><td>22</td></tr>

<tr><td>UPDATE</td><td>23</td></tr>

<tr><td>LABEL</td><td>24</td></tr>

<tr><td>MERGE</td><td>25</td></tr>

<tr><td>BASELINE_CONTROL</td><td>26</td></tr>

<tr><td>MKACTIVITY</td><td>27</td></tr>

</table>

<p>Later version of ajp13, will transport

additional methods, even if they are not in this list.</p>

</section>

<section><title>protocol, req_uri, remote_addr, remote_host, server_name,

server_port, is_ssl</title>

<p>These are all fairly self-explanatory. Each of these is required, and

will be sent for every request.</p>

</section>

<section><title>Headers</title>

<p>The structure of <code>request_headers</code> is the following:

First, the number of headers <code>num_headers</code> is encoded.

Then, a series of header name <code>req_header_name</code> / value

<code>req_header_value</code> pairs follows.

Common header names are encoded as integers,

to save space. If the header name is not in the list of basic headers,

it is encoded normally (as a string, with prefixed length). The list of

common headers <code>sc_req_header_name</code>and their codes

is as follows (all are case-sensitive):</p>

<table>

<tr><td>Name</td><td>Code value</td><td>Code name</td></tr>

<tr><td>accept</td><td>0xA001</td><td>SC_REQ_ACCEPT</td></tr>

<tr><td>accept-charset</td><td>0xA002</td><td>SC_REQ_ACCEPT_CHARSET

</td></tr>

<tr><td>accept-encoding</td><td>0xA003</td><td>SC_REQ_ACCEPT_ENCODING

</td></tr>

<tr><td>accept-language</td><td>0xA004</td><td>SC_REQ_ACCEPT_LANGUAGE

</td></tr>

<tr><td>authorization</td><td>0xA005</td><td>SC_REQ_AUTHORIZATION</td>

</tr>

<tr><td>connection</td><td>0xA006</td><td>SC_REQ_CONNECTION</td></tr>

<tr><td>content-type</td><td>0xA007</td><td>SC_REQ_CONTENT_TYPE</td>

</tr>

<tr><td>content-length</td><td>0xA008</td><td>SC_REQ_CONTENT_LENGTH</td>

</tr>

<tr><td>cookie</td><td>0xA009</td><td>SC_REQ_COOKIE</td></tr>

<tr><td>cookie2</td><td>0xA00A</td><td>SC_REQ_COOKIE2</td></tr>

<tr><td>host</td><td>0xA00B</td><td>SC_REQ_HOST</td></tr>

<tr><td>pragma</td><td>0xA00C</td><td>SC_REQ_PRAGMA</td></tr>

<tr><td>referer</td><td>0xA00D</td><td>SC_REQ_REFERER</td></tr>

<tr><td>user-agent</td><td>0xA00E</td><td>SC_REQ_USER_AGENT</td></tr>

</table>

<p>The Java code that reads this grabs the first two-byte integer and if

it sees an <code>'0xA0'</code> in the most significant

byte, it uses the integer in the second byte as an index into an array of

header names. If the first byte is not <code>0xA0</code>, it assumes that

the two-byte integer is the length of a string, which is then read in.</p>

<p>This works on the assumption that no header names will have length

greater than <code>0x9999 (==0xA000 - 1)</code>, which is perfectly

reasonable, though somewhat arbitrary.</p>

<note><title>Note:</title>

The <code>content-length</code> header is extremely

important. If it is present and non-zero, the container assumes that

the request has a body (a POST request, for example), and immediately

reads a separate packet off the input stream to get that body.

</note>

</section>

<section><title>Attributes</title>

<p>The attributes prefixed with a <code>?</code>

(e.g. <code>?context</code>) are all optional. For each, there is a

single byte code to indicate the type of attribute, and then a string to

give its value. They can be sent in any order (thogh the C code always

sends them in the order listed below). A special terminating code is

sent to signal the end of the list of optional attributes. The list of

byte codes is:</p>

<table>

<tr><td>Information</td><td>Code Value</td><td>Note</td></tr>

<tr><td>?context</td><td>0x01</td><td>Not currently implemented

</td></tr>

<tr><td>?servlet_path</td><td>0x02</td><td>Not currently implemented

</td></tr>

<tr><td>?remote_user</td><td>0x03</td><td></td></tr>

<tr><td>?auth_type</td><td>0x04</td><td></td></tr>

<tr><td>?query_string</td><td>0x05</td><td></td></tr>

<tr><td>?jvm_route</td><td>0x06</td><td></td></tr>

<tr><td>?ssl_cert</td><td>0x07</td><td></td></tr>

<tr><td>?ssl_cipher</td><td>0x08</td><td></td></tr>

<tr><td>?ssl_session</td><td>0x09</td><td></td></tr>

<tr><td>?req_attribute</td><td>0x0A</td><td>Name (the name of the

attribute follows)</td></tr>

<tr><td>?ssl_key_size</td><td>0x0B</td><td></td></tr>

<tr><td>are_done</td><td>0xFF</td><td>request_terminator</td></tr>

</table>

<p>The <code>context</code> and <code>servlet_path</code> are not

currently set by the C code, and most of the Java code completely ignores

whatever is sent over for those fields (and some of it will actually break

if a string is sent along after one of those codes). I don't know if this

is a bug or an unimplemented feature or just vestigial code, but it's

missing from both sides of the connection.</p>

<p>The <code>remote_user</code> and <code>auth_type</code> presumably

refer to HTTP-level authentication, and communicate the remote user's

username and the type of authentication used to establish their identity

(e.g. Basic, Digest).</p>

<p>The <code>query_string</code>, <code>ssl_cert</code>,

<code>ssl_cipher</code>, and <code>ssl_session</code> refer to the

corresponding pieces of HTTP and HTTPS.</p>

<p>The <code>jvm_route</code>, is used to support sticky

sessions -- associating a user's sesson with a particular Tomcat instance

in the presence of multiple, load-balancing servers.</p>

<p>Beyond this list of basic attributes, any number of other attributes

can be sent via the <code>req_attribute</code> code <code>0x0A</code>.

A pair of strings to represent the attribute name and value are sent

immediately after each instance of that code. Environment values are passed

in via this method.</p>

<p>Finally, after all the attributes have been sent, the attribute

terminator, <code>0xFF</code>, is sent. This signals both the end of the

list of attributes and also then end of the Request Packet.</p>

</section>

</section>

<section id="resppacketstruct"><title>Response Packet Structure</title>

<p>for messages which the container can send back to the server.</p>

<example><pre>

AJP13_SEND_BODY_CHUNK :=

prefix_code 3

chunk_length (integer)

chunk *(byte)

AJP13_SEND_HEADERS :=

prefix_code 4

http_status_code (integer)

http_status_msg (string)

num_headers (integer)

response_headers *(res_header_name header_value)

res_header_name :=

sc_res_header_name | (string) [see below for how this is parsed]

sc_res_header_name := 0xA0 (byte)

header_value := (string)

AJP13_END_RESPONSE :=

prefix_code 5

reuse (boolean)

AJP13_GET_BODY_CHUNK :=

prefix_code 6

requested_length (integer)

</pre></example>

<section><title>Details:</title></section>

<section><title>Send Body Chunk</title>

<p>The chunk is basically binary data, and is sent directly back to the

browser.</p>

</section>

<section><title>Send Headers</title>

<p>The status code and message are the usual HTTP things

(e.g. <code>200</code> and <code>OK</code>). The response header names are

encoded the same way the request header names are. See header_encoding above

for details about how the the codes are distinguished from the strings.<br />

The codes for common headers are:</p>

<table>

<tr><td>Name</td><td>Code value</td></tr>

<tr><td>Content-Type</td><td>0xA001</td></tr>

<tr><td>Content-Language</td><td>0xA002</td></tr>

<tr><td>Content-Length</td><td>0xA003</td></tr>

<tr><td>Date</td><td>0xA004</td></tr>

<tr><td>Last-Modified</td><td>0xA005</td></tr>

<tr><td>Location</td><td>0xA006</td></tr>

<tr><td>Set-Cookie</td><td>0xA007</td></tr>

<tr><td>Set-Cookie2</td><td>0xA008</td></tr>

<tr><td>Servlet-Engine</td><td>0xA009</td></tr>

<tr><td>Status</td><td>0xA00A</td></tr>

<tr><td>WWW-Authenticate</td><td>0xA00B</td></tr>

</table>

<p> After the code or the string header name, the header value is

immediately encoded.</p>

</section>

<section><title>End Response</title>

<p>Signals the end of this request-handling cycle. If the

<code>reuse</code> flag is true <code>(==1)</code>, this TCP connection can

now be used to handle new incoming requests. If <code>reuse</code> is false

(anything other than 1 in the actual C code), the connection should

be closed.</p>

</section>

<section><title>Get Body Chunk</title>

<p>The container asks for more data from the request (If the body was

too large to fit in the first packet sent over or when the request is

chuncked). The server will send a body packet back with an amount of data

which is the minimum of the <code>request_length</code>, the maximum send

body size <code>(8186 (8 Kbytes - 6))</code>, and the number of bytes

actually left to send from the request body.<br/>

If there is no more data in the body (i.e. the servlet container is

trying to read past the end of the body), the server will send back an

<em>empty</em> packet, which is a body packet with a payload length of 0.

<code>(0x12,0x34,0x00,0x00)</code></p>

</section>

</section>

</modulesynopsis>