mod_cgi.html revision 1319d315ae7b55c02ea1cfb4c6843c04059bca76
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<title>Apache module mod_cgi</title>
</head>
<!-- Background white, links blue (unvisited), navy (visited), red (active) -->
<body bgcolor="#FFFFFF" text="#000000" link="#0000FF"
vlink="#000080" alink="#FF0000">
<!--#include virtual="header.html" -->
<h1 align="CENTER">Module mod_cgi</h1>
<p>This module provides for execution of CGI scripts.</p>
rel="Help"><strong>Status:</strong></a> Base<br />
<a href="module-dict.html#SourceFile"
<a href="module-dict.html#ModuleIdentifier"
rel="Help"><strong>Module Identifier:</strong></a>
cgi_module</p>
<h2>Summary</h2>
<!-- XXX: Should have references to CGI definition/RFC -->
<!-- XXX: Should mention Options ExecCGI -->
<!-- XXX: Should mention AcceptPathInfo -->
<p>Any file that has the mime type
<code>cgi-script</code> (Apache 1.1 or later) will be treated
as a CGI script, and run by the server, with its output being
returned to the client. Files acquire this type either by
having a name containing an extension defined by the <a
href="mod_mime.html#addtype">AddType</a> directive, or by being
directory.</p>
<p>When the server invokes a CGI script, it will add a variable
called <code>DOCUMENT_ROOT</code> to the environment. This
variable will contain the value of the <a
href="core.html#documentroot">DocumentRoot</a> configuration
variable.</p>
<p>For an introduction to using CGI scripts with Apache, see
With CGI</a>.</p>
<p>When using a multi-threaded MPM under unix, the module <a
href="mod_cgid.html">mod_cgid</a> should be used in place of
this module. At the user level, the two modules are essentially
identical.</p>
<h2>Directives</h2>
<ul>
<li><a href="#scriptlog">ScriptLog</a></li>
<li><a href="#scriptloglength">ScriptLogLength</a></li>
<li><a href="#scriptlogbuffer">ScriptLogBuffer</a></li>
</ul>
<h2>CGI Environment variables</h2>
The server will set the CGI environment variables as described
in the <a href="http://hoohoo.ncsa.uiuc.edu/cgi/">CGI
specification</a>, with the following provisions:
<dl>
<dt>PATH_INFO</dt>
<dd>This will not be available if the <a
directive is explicitly set to <code>off</code>. The default
behavior, if AcceptPathInfo is not given, is that mod_cgi will
filename in the URI), while the core server will return a 404
NOT FOUND error for requests with additional path info.
Omitting the AcceptPathInfo directive has the same effect as
setting it <code>on</code> for mod_cgi requests.</dd>
<dt>REMOTE_HOST</dt>
<dd>This will only be set if <a
is set to <code>on</code> (it is off by default), and if a
reverse DNS lookup of the accessing host's address indeed
finds a host name.</dd>
<dt>REMOTE_IDENT</dt>
<dd>This will only be set if <a
href="core.html#identitycheck">IdentityCheck</a> is set to
<code>on</code> and the accessing host supports the ident
protocol. Note that the contents of this variable cannot be
relied upon because it can easily be faked, and if there is a
proxy between the client and the server, it is usually
totally useless.</dd>
<dt>REMOTE_USER</dt>
<dd>This will only be set if the CGI script is subject to
authentication.</dd>
</dl>
<h2><a id="cgi_debug" name="cgi_debug">CGI Debugging</a></h2>
Debugging CGI scripts has traditionally been difficult, mainly
because it has not been possible to study the output (standard
output and error) for scripts which are failing to run
properly. These directives, included in Apache 1.2 and later,
provide more detailed logging of errors when they occur.
<h2>CGI Logfile Format</h2>
When configured, the CGI error log logs any CGI which does not
execute properly. Each CGI script which fails to operate causes
several lines of information to be logged. The first two lines
are always of the format:
<pre>
%% [<em>time</em>] <em>request-line</em>
%% <em>HTTP-status</em> <em>CGI-script-filename</em>
</pre>
If the error is that CGI script cannot be run, the log file
will contain an extra two lines:
<pre>
%%error
<em>error-message</em>
</pre>
Alternatively, if the error is the result of the script
returning incorrect header information (often due to a bug in
the script), the following information is logged:
<pre>
%request
<em>All HTTP request headers received</em>
<em>POST or PUT entity (if any)</em>
%response
<em>All headers output by the CGI script</em>
%stdout
<em>CGI standard output</em>
%stderr
<em>CGI standard error</em>
</pre>
(The %stdout and %stderr parts may be missing if the script did
not output anything on standard output or standard error).
<hr />
<h3><a id="scriptlog" name="scriptlog">ScriptLog</a>
directive</h3>
<a href="directive-dict.html#Syntax"
rel="Help"><strong>Syntax:</strong></a> ScriptLog
<em>filename</em><br />
<a href="directive-dict.html#Default"
rel="Help"><strong>Default:</strong></a> none<br />
<a href="directive-dict.html#Context"
rel="Help"><strong>Context:</strong></a> server config<br />
<a href="directive-dict.html#Status"
rel="Help"><strong>Status:</strong></a> mod_cgi
<p>The <tt>ScriptLog</tt> directive sets the CGI script error
logfile. If no ScriptLog is given, no error log is created. If
given, any CGI errors are logged into the filename given as
argument. If this is a relative file or path it is taken
relative to the server root.</p>
<p>This log will be opened as the user the child processes run
as, ie. the user specified in the main <a
href="core.html#User">User</a> directive. This means that
either the directory the script log is in needs to be writable
by that user or the file needs to be manually created and set
to be writable by that user. If you place the script log in
your main logs directory, do <strong>NOT</strong> change the
directory permissions to make it writable by the user the child
processes run as.</p>
<p>Note that script logging is meant to be a debugging feature
when writing CGI scripts, and is not meant to be activated
continuously on running servers. It is not optimized for speed
or efficiency, and may have security problems if used in a
manner other than that for which it was designed.</p>
<hr />
<h3><a id="scriptloglength"
name="scriptloglength">ScriptLogLength</a> directive</h3>
<a href="directive-dict.html#Syntax"
rel="Help"><strong>Syntax:</strong></a> ScriptLogLength
<em>bytes</em><br />
<a href="directive-dict.html#Default"
rel="Help"><strong>Default:</strong></a> 10385760<br />
<a href="directive-dict.html#Context"
rel="Help"><strong>Context:</strong></a> server config<br />
<a href="directive-dict.html#Status"
rel="Help"><strong>Status:</strong></a> mod_cgi
<p><tt>ScriptLogLength</tt> can be used to limit the size of
the CGI script logfile. Since the logfile logs a lot of
information per CGI error (all request headers, all script
output) it can grow to be a big file. To prevent problems due
to unbounded growth, this directive can be used to set an
maximum file-size for the CGI logfile. If the file exceeds this
size, no more information will be written to it.</p>
<hr />
<h3><a id="scriptlogbuffer"
name="scriptlogbuffer">ScriptLogBuffer</a></h3>
<a href="directive-dict.html#Syntax"
rel="Help"><strong>Syntax:</strong></a> ScriptLogBuffer
<em>bytes</em><br />
<a href="directive-dict.html#Default"
rel="Help"><strong>Default:</strong></a> 1024<br />
<a href="directive-dict.html#Context"
rel="Help"><strong>Context:</strong></a> server config<br />
<a href="directive-dict.html#Status"
rel="Help"><strong>Status:</strong></a> mod_cgi
<p>The size of any PUT or POST entity body that is logged to
the file is limited, to prevent the log file growing too big
too quickly if large bodies are being received. By default, up
to 1024 bytes are logged, but this can be changed with this
directive. <!--#include virtual="footer.html" -->
</p>
</body>
</html>