File: README.html

package info (click to toggle)
speedy-cgi-perl 2.22-1
links: PTS
area: main
in suites: sarge
size: 1,048 kB
ctags: 886
sloc: ansic: 4,487; perl: 945; sh: 883; makefile: 75
file content (1142 lines) | stat: -rw-r--r-- 34,985 bytes
parent folder | download | duplicates (4)
<HTML>
<HEAD>
<TITLE>SpeedyCGI - Speed up perl scripts by running them persistently.</TITLE>
<LINK REV="made" HREF="mailto:root@porky.devel.redhat.com">
</HEAD>

<BODY>

<!-- INDEX BEGIN -->

<UL>

	<LI><A HREF="#NAME">NAME</A>
	<LI><A HREF="#SYNOPSIS">SYNOPSIS</A>
	<LI><A HREF="#DESCRIPTION">DESCRIPTION</A>
	<LI><A HREF="#OPTIONS">OPTIONS</A>
	<UL>

		<LI><A HREF="#Setting_Option_Values">Setting Option Values</A>
		<LI><A HREF="#Context">Context</A>
		<LI><A HREF="#Options_Available">Options Available</A>
	</UL>

	<LI><A HREF="#METHODS">METHODS</A>
	<LI><A HREF="#INSTALLATION">INSTALLATION</A>
	<UL>

		<LI><A HREF="#Binary_Installation">Binary Installation</A>
		<LI><A HREF="#Source_Code_Installation">Source Code Installation</A>
		<LI><A HREF="#Standard_Install">Standard Install</A>
		<LI><A HREF="#Install_in_a_Different_Directory">Install in a Different Directory</A>
		<LI><A HREF="#Setuid_Install">Setuid Install</A>
		<LI><A HREF="#Apache_Installation">Apache Installation</A>
		<LI><A HREF="#Apache_Configuration">Apache Configuration</A>
	</UL>

	<LI><A HREF="#FREQUENTLY_ASKED_QUESTIONS">FREQUENTLY ASKED QUESTIONS</A>
	<LI><A HREF="#USING_GROUPS">USING GROUPS</A>
	<LI><A HREF="#DOWNLOADING">DOWNLOADING</A>
	<UL>

		<LI><A HREF="#Binaries">Binaries</A>
		<LI><A HREF="#Source_Code">Source Code</A>
	</UL>

	<LI><A HREF="#AUTHOR">AUTHOR</A>
	<UL>

		<LI><A HREF="#Contributors">Contributors</A>
	</UL>

	<LI><A HREF="#SEE_ALSO">SEE ALSO</A>
	<LI><A HREF="#MORE_INFORMATION">MORE INFORMATION</A>
	<UL>

		<LI><A HREF="#SpeedyCGI_Home_Page">SpeedyCGI Home Page</A>
		<LI><A HREF="#Mailing_List">Mailing List</A>
		<LI><A HREF="#Bugs_and_Todo_List">Bugs and Todo List</A>
		<LI><A HREF="#Japanese_Translation">Japanese Translation</A>
		<LI><A HREF="#Benchmarks">Benchmarks</A>
		<LI><A HREF="#Success_Stories">Success Stories</A>
		<LI><A HREF="#Revision_History">Revision History</A>
		<LI><A HREF="#YAPC_2001_Presentation">YAPC 2001 Presentation</A>
	</UL>

	<LI><A HREF="#COPYRIGHT">COPYRIGHT</A>
</UL>
<!-- INDEX END -->

<HR>
<P>
<H1><A NAME="NAME">NAME</A></H1>
<P>
SpeedyCGI - Speed up perl scripts by running them persistently.

<P>
<HR>
<H1><A NAME="SYNOPSIS">SYNOPSIS</A></H1>
<P>
<PRE> #!/usr/bin/speedy
</PRE>
<P>
<PRE> ### Your Script Here.  For example:
 print &quot;Content-type: text/html\n\nHello World!\n&quot;;
</PRE>
<P>
<PRE> ##
 ## Optionally, use the CGI::SpeedyCGI module for various things
 ##
</PRE>
<P>
<PRE> # Create a SpeedyCGI object
 use CGI::SpeedyCGI;
 my $sp = CGI::SpeedyCGI-&gt;new;
</PRE>
<P>
<PRE> # See if we are running under SpeedyCGI or not.
 print &quot;Running under speedy=&quot;, $sp-&gt;i_am_speedy ? 'yes' : 'no', &quot;\n&quot;;
</PRE>
<P>
<PRE> # Register a shutdown handler
 $sp-&gt;add_shutdown_handler(sub { do something here });
</PRE>
<P>
<PRE> # Register a cleanup handler
 $sp-&gt;register_cleanup(sub { do something here });
</PRE>
<P>
<PRE> # Set/get some SpeedyCGI options
 $sp-&gt;setopt('timeout', 30);
 print &quot;maxruns=&quot;, $sp-&gt;getopt('maxruns'), &quot;\n&quot;;
</PRE>
<P>
<HR>
<H1><A NAME="DESCRIPTION">DESCRIPTION</A></H1>
<P>
SpeedyCGI is a way to run perl scripts persistently, which can make them
run much more quickly. A script can be made to to run persistently by
changing the interpreter line at the top of the script from:

<P>
<PRE>    #!/usr/bin/perl
</PRE>
<P>
to

<P>
<PRE>    #!/usr/bin/speedy
</PRE>
<P>
After the script is initially run, instead of exiting, the perl interpreter
is kept running. During subsequent runs, this interpreter is used to handle
new executions instead of starting a new perl interpreter each time. A very
fast frontend program, written in C, is executed for each request. This
fast frontend then contacts the persistent Perl process, which is usually
already running, to do the work and return the results.

<P>
By default each perl script runs in its own Unix process, so one perl
script can't interfere with another. Command line options can also be used
to deal with programs that have memory leaks or other problems that might
keep them from otherwise running persistently.

<P>
SpeedyCGI can be used to speed up perl CGI scripts. It conforms to the CGI
specification, and does not run perl code inside the web server. Since the
perl interpreter runs outside the web server, it can't cause problems for
the web server itself.

<P>
SpeedyCGI also provides an Apache module so that under the Apache web
server, scripts can be run without the overhead of doing a fork/exec for
each request. With this module a small amount of frontend code is run
within the web server - the perl interpreters still run outside the server.

<P>
SpeedyCGI and PersistentPerl are currently both names for the same code.
SpeedyCGI was the original name, but because people weren't sure what it
did, the name PersistentPerl was picked as an alias. At some point
SpeedyCGI will be replaced by PersistentPerl, or become a sub-class of
PersistentPerl to avoid always having two distributions.

<P>
<HR>
<H1><A NAME="OPTIONS">OPTIONS</A></H1>
<P>
<HR>
<H2><A NAME="Setting_Option_Values">Setting Option Values</A></H2>
<P>
SpeedyCGI options can be set in several ways:

<DL>
<DT><STRONG><A NAME="item_Command">Command Line</A></STRONG><DD>
<P>
The speedy command line is the same as for regular perl, with the exception
that SpeedyCGI specific options can be passed in after a ``--''.

<P>
For example the line:

<P>
<PRE>        #!/usr/bin/speedy -w -- -t300
</PRE>
<P>
at the top of your script will set the perl option ``<CODE>-w</CODE>'' and will pass the ``<CODE>-t</CODE>'' option to SpeedyCGI, setting the Timeout value to 300 seconds.

</DL>
<DL>
<DT><STRONG><A NAME="item_Environment">Environment</A></STRONG><DD>
<P>
Environment variables can be used to pass in options. This can only be done
before the initial execution, not from within the script itself. The name
of the environment variable is always SPEEDY_ followed by the option name
in upper-case. For example to set the speedy Timeout option, use the
environment variable named SPEEDY_TIMEOUT.

</DL>
<DL>
<DT><STRONG><A NAME="item_Module">Module</A></STRONG><DD>
<P>
The CGI::SpeedyCGI module provides the setopt method to set options from
within the perl script at runtime. There is also a getopt method to
retrieve the current options. See <A HREF="#METHODS">METHODS</A> below.

</DL>
<DL>
<DT><STRONG><A NAME="item_Apache">Apache</A></STRONG><DD>
<P>
If you are using the optional Apache module, SpeedyCGI options can be set
in the <EM>httpd.conf</EM> file. The name of the apache directive will always be Speedy followed by
the option name. For example to set the Timeout option, use the apache
directive SpeedyTimeout.

</DL>
<P>
<HR>
<H2><A NAME="Context">Context</A></H2>
<P>
Not all options below are available in all contexts. The context for which
each option is valid is listed on the ``Context'' line in the section
below. There are three contexts:

<DL>
<DT><STRONG><A NAME="item_speedy">speedy</A></STRONG><DD>
<P>
The command-line ``speedy'' program, used normally with #! at the top of
your script or from a shell prompt.

</DL>
<DL>
<DT><STRONG><A NAME="item_mod_speedycgi">mod_speedycgi</A></STRONG><DD>
<P>
The optional Apache mod_speedycgi module.

</DL>
<DL>
<DT><STRONG><A NAME="item_module">module</A></STRONG><DD>
<P>
During perl execution via the CGI::SpeedyCGI module's getopt/setopt
methods.

</DL>
<P>
<HR>
<H2><A NAME="Options_Available">Options Available</A></H2>
<DL>
<DT><STRONG><A NAME="item_BackendProg">BackendProg</A></STRONG><DD>
<P>
<PRE>    Command Line    : -p&lt;string&gt;
    Default Value   : &quot;/usr/bin/speedy_backend&quot;
    Context         : mod_speedycgi, speedy
</PRE>
<P>
<PRE>    Description:
</PRE>
<P>
<PRE>        Path to the speedy backend program.
</PRE>
<DT><STRONG><A NAME="item_BufsizGet">BufsizGet</A></STRONG><DD>
<P>
<PRE>    Command Line    : -B&lt;number&gt;
    Default Value   : 131072
    Context         : speedy
</PRE>
<P>
<PRE>    Description:
</PRE>
<P>
<PRE>        Use &lt;number&gt; bytes as the maximum size for the buffer that
        receives data from the perl backend.
</PRE>
<DT><STRONG><A NAME="item_BufsizPost">BufsizPost</A></STRONG><DD>
<P>
<PRE>    Command Line    : -b&lt;number&gt;
    Default Value   : 131072
    Context         : speedy
</PRE>
<P>
<PRE>    Description:
</PRE>
<P>
<PRE>        Use &lt;number&gt; bytes as the maximum size for the buffer that
        sends data to the perl backend.
</PRE>
<DT><STRONG><A NAME="item_Group">Group</A></STRONG><DD>
<P>
<PRE>    Command Line    : -g&lt;string&gt;
    Default Value   : &quot;none&quot;
    Context         : mod_speedycgi, speedy
</PRE>
<P>
<PRE>    Description:
</PRE>
<P>
<PRE>        Allow a single perl interpreter to run multiple scripts.
        All scripts that are run with the same group name and by
        the same user will be run by the same group of perl
        interpreters. If the group name is &quot;none&quot; then grouping is
        disabled and each interpreter will run one script.
        Different group names allow scripts to be separated into
        different groups. Name is case-sensitive, and only the
        first 12-characters are significant. Specifying an empty
        group name is the same as specifying the group name
        &quot;default&quot; - this allows just specifying &quot;-g&quot; on the command
        line to turn on grouping.
</PRE>
<DT><STRONG><A NAME="item_MaxBackends">MaxBackends</A></STRONG><DD>
<P>
<PRE>    Command Line    : -M&lt;number&gt;
    Default Value   : 0 (no max)
    Context         : mod_speedycgi, speedy
</PRE>
<P>
<PRE>    Description:
</PRE>
<P>
<PRE>        If non-zero, limits the number of speedy backends running
        for this perl script to &lt;number&gt;.
</PRE>
<DT><STRONG><A NAME="item_MaxRuns">MaxRuns</A></STRONG><DD>
<P>
<PRE>    Command Line    : -r&lt;number&gt;
    Default Value   : 500
    Context         : mod_speedycgi, module, speedy
</PRE>
<P>
<PRE>    Description:
</PRE>
<P>
<PRE>        Once the perl interpreter has run &lt;number&gt; times, re-exec
        the backend process.  Zero indicates no maximum.  This
        option is useful for processes that tend to consume
        resources over time.
</PRE>
<DT><STRONG><A NAME="item_PerlArgs">PerlArgs</A></STRONG><DD>
<P>
<PRE>    Command Line    : N/A
    Default Value   : &quot;&quot;
    Context         : mod_speedycgi
</PRE>
<P>
<PRE>    Description:
</PRE>
<P>
<PRE>        Command-line options to pass to the perl interpreter.
</PRE>
<DT><STRONG><A NAME="item_Timeout">Timeout</A></STRONG><DD>
<P>
<PRE>    Command Line    : -t&lt;number&gt;
    Default Value   : 3600 (one hour)
    Context         : mod_speedycgi, module, speedy
</PRE>
<P>
<PRE>    Description:
</PRE>
<P>
<PRE>        If no new requests have been received after &lt;number&gt;
        seconds, exit the persistent perl interpreter.  Zero
        indicates no timeout.
</PRE>
<DT><STRONG><A NAME="item_TmpBase">TmpBase</A></STRONG><DD>
<P>
<PRE>    Command Line    : -T&lt;string&gt;
    Default Value   : &quot;/tmp/speedy&quot;
    Context         : mod_speedycgi, speedy
</PRE>
<P>
<PRE>    Description:
</PRE>
<P>
<PRE>        Use the given prefix for creating temporary files.  This
        must be a filename prefix, not a directory name.
</PRE>
<DT><STRONG><A NAME="item_Version">Version</A></STRONG><DD>
<P>
<PRE>    Command Line    : -v
    Context         : speedy
</PRE>
<P>
<PRE>    Description:
</PRE>
<P>
<PRE>        Print the SpeedyCGI version and exit.
</PRE>
</DL>
<P>
<HR>
<H1><A NAME="METHODS">METHODS</A></H1>
<P>
The following methods are available in the CGI::SpeedyCGI module.

<DL>
<DT><STRONG><A NAME="item_new">new</A></STRONG><DD>
<P>
Create a new CGI::SpeedyCGI object.

<P>
<PRE>    my $sp = CGI::SpeedyCGI-&gt;new;
</PRE>
<DT><STRONG><A NAME="item_register_cleanup">register_cleanup($function_ref)</A></STRONG><DD>
<P>
Register a function that will be called at the end of each request, after
your script finishes running, but before STDOUT and STDERR are closed.
Multiple functions can be added by calling the method more than once. At
the end of the request, each function will be called in the order in which
it was registered.

<P>
<PRE>    $sp-&gt;register_cleanup(\&amp;cleanup_func);
</PRE>
<DT><STRONG><A NAME="item_add_shutdown_handler">add_shutdown_handler($function_ref)</A></STRONG><DD>
<P>
Add a function to the list of functions that will be called right before
the perl interpreter exits. This is <STRONG>not</STRONG> at the end of each request, it is when the perl interpreter decides to exit
completely due to a Timeout or reaching MaxRuns.

<P>
<PRE>    $sp-&gt;add_shutdown_handler(sub {$dbh-&gt;logout});
</PRE>
<DT><STRONG><A NAME="item_set_shutdown_handler">set_shutdown_handler($function_ref)</A></STRONG><DD>
<P>
Deprecated. Similar to <A HREF="#item_add_shutdown_handler">add_shutdown_handler</A>, but only allows for a single function to be registered.

<P>
<PRE>    $sp-&gt;set_shutdown_handler(sub {$dbh-&gt;logout});
</PRE>
<DT><STRONG><A NAME="item_i_am_speedy">i_am_speedy</A></STRONG><DD>
<P>
Returns a boolean telling whether this script is running under SpeedyCGI or
not. A perl script can run under regular perl, or under SpeedyCGI. This
method allows the script to tell which environment it is in.

<P>
<PRE>    $sp-&gt;i_am_speedy;
</PRE>
<P>
To make your script as portable as possible, you can use the following test
to make sure both the SpeedyCGI module is available and you are running
under SpeedyCGI:

<P>
<PRE>    if (eval {require CGI::SpeedyCGI} &amp;&amp; CGI::SpeedyCGI-&gt;i_am_speedy) {
        Do something SpeedyCGI specific here...
</PRE>
<P>
To increase the speed of this check you can also test whether the following
variable is defined instead of going through the object interface:

<P>
<PRE>    $CGI::SpeedyCGI::i_am_speedy
</PRE>
<DT><STRONG><A NAME="item_setopt">setopt($optname, $value)</A></STRONG><DD>
<P>
Set one of the SpeedyCGI options given in <A HREF="#Options_Available">Options Available</A>. Returns the option's previous value. <CODE>$optname</CODE> is
case-insensitive.

<P>
<PRE>    $sp-&gt;setopt('TIMEOUT', 300);
</PRE>
<DT><STRONG><A NAME="item_getopt">getopt($optname)</A></STRONG><DD>
<P>
Return the current value of one of the SpeedyCGI options.
<CODE>$optname</CODE> is case-insensitive.

<P>
<PRE>    $sp-&gt;getopt('TIMEOUT');
</PRE>
<DT><STRONG><A NAME="item_shutdown_now">shutdown_now</A></STRONG><DD>
<P>
Shut down the perl interpreter right away. This function does not return.

<P>
<PRE>    $sp-&gt;shutdown_now
</PRE>
<DT><STRONG><A NAME="item_shutdown_next_time">shutdown_next_time</A></STRONG><DD>
<P>
Shut down the perl interpreter as soon as this request is done.

<P>
<PRE>    $sp-&gt;shutdown_next_time
</PRE>
</DL>
<P>
<HR>
<H1><A NAME="INSTALLATION">INSTALLATION</A></H1>
<P>
To install SpeedyCGI you will need to either download a binary package for
your OS, or compile SpeedyCGI from source code. See <A HREF="#DOWNLOADING">DOWNLOADING</A>
for information on where to obtain the source code and binaries.

<P>
<HR>
<H2><A NAME="Binary_Installation">Binary Installation</A></H2>
<P>
Once you have downloaded the binary package for your OS, you'll need to
install it using the normal package tools for your OS. The commands to do
that are:

<DL>
<DT><STRONG><A NAME="item_Linux">Linux</A></STRONG><DD>
<P>
<PRE> rpm -i &lt;filename&gt;
</PRE>
<DT><STRONG><A NAME="item_Solaris">Solaris</A></STRONG><DD>
<P>
<PRE> gunzip &lt;filename&gt;.gz
 pkgadd -d &lt;filename&gt;
</PRE>
<DT><STRONG><A NAME="item_BSD">BSD</A></STRONG><DD>
<P>
<PRE> pkg_add &lt;filename&gt;
</PRE>
</DL>
<P>
If you are also installing the apache module you will have to configure
Apache as documented in <A HREF="#Apache_Configuration">Apache Configuration</A>.

<P>
<HR>
<H2><A NAME="Source_Code_Installation">Source Code Installation</A></H2>
<P>
To compile SpeedyCGI you will need perl 5.005_03 or later, and a C
compiler, preferably the same one that your perl distribution was compiled
with. SpeedyCGI is known to work under Solaris, Redhat Linux, FreeBSD and
OpenBSD. There may be problems with other OSes or earlier versions of Perl.
SpeedyCGI may not work with threaded perl -- as of release 2.10, Linux and
Solaris seem to work OK with threaded perl, but FreeBSD does not.

<P>
<HR>
<H2><A NAME="Standard_Install">Standard Install</A></H2>
<P>
To do a standard install from source code, execute the following:

<P>
<PRE>    perl Makefile.PL
    make
    make test
    make install
</PRE>
<P>
This will install the speedy and speedy_backend binaries in the same
directory where perl was installed, and the SpeedyCGI.pm module in the
standard perl lib directory. It will also attempt to install the
mod_speedycgi module if you have the command <STRONG>apxs</STRONG> in your path.

<P>
<HR>
<H2><A NAME="Install_in_a_Different_Directory">Install in a Different Directory</A></H2>
<P>
If you don't have permission to install into the standard perl directory,
or if you want to install elsewhere, the easiest way is to compile and
install your own copy of perl in another location, then use your new
version of perl when you run ``perl Makefile.PL''. The SpeedyCGI binaries
and module will then be installed in the same location as the new version
of perl.

<P>
If you can't install your own perl, you can take the following steps:

<UL>
<LI>
<P>
Edit src/optdefs and change the default value for BackendProg to the
location where speedy_backend will be installed.

</UL>
<UL>
<LI>
<P>
Compile as above, then manually copy the speedy and speedy_backend binaries
to where you want to install them.

</UL>
<UL>
<LI>
<P>
If you want to use the CGI::SpeedyCGI module in your code (it's not
required), you will have to use ``use lib'' so it can be located.

</UL>
<P>
<HR>
<H2><A NAME="Setuid_Install">Setuid Install</A></H2>
<P>
SpeedyCGI has limited support for running setuid - installing this way may
compromise the security of your system. To install setuid do the following:

<UL>
<LI>
<P>
Run ``perl Makefile.PL''

</UL>
<UL>
<LI>
<P>
Edit speedy/Makefile and add ``-DIAMSUID'' to the end of the ``DEFINE = ''
line.

</UL>
<UL>
<LI>
<P>
Run make

</UL>
<UL>
<LI>
<P>
Take the resulting ``speedy'' binary and install it suid-root as
/usr/bin/speedy_suid

</UL>
<UL>
<LI>
<P>
Change your setuid scripts to use /usr/bin/speedy_suid as the interpreter.

</UL>
<P>
This has been know to work in Linux and FreeBSD. Solaris will work as long
as the Group option is set to ``none''.

<P>
<HR>
<H2><A NAME="Apache_Installation">Apache Installation</A></H2>
<P>
To compile the optional apache mod_speedycgi module you must have the <STRONG>apxs</STRONG>
command in your path. Redhat includes this command with the
``apache-devel'' RPM, though it may not work properly for installation.

<P>
If the apache installation fails:

<UL>
<LI>
<P>
Copy the mod_speedycgi.so from the mod_speedycgi directory, or from the
mod_speedycgi2/.libs directory, to wherever your apache modules are stored
(try <EM>/usr/lib/apache</EM>)

</UL>
<UL>
<LI>
<P>
Edit your <EM>httpd.conf</EM> (try <EM>/etc/httpd/conf/httpd.conf</EM>) and add the following lines. The path at the end of the LoadModule
directive may be different in your installation -- look at other
LoadModules to see.

<P>
<PRE>    LoadModule speedycgi_module modules/mod_speedycgi.so
</PRE>
<P>
If you are using Apache-1, also add:

<P>
<PRE>    AddModule mod_speedycgi.c
</PRE>
</UL>
<P>
<HR>
<H2><A NAME="Apache_Configuration">Apache Configuration</A></H2>
<P>
Once mod_speedycgi is installed, it has to be configured to be used for
your perl scripts. There are two methods.

<P>
Warning! The instructions below may compromise the security of your web
site. The security risks associated with SpeedyCGI are similar to those of
regular CGI. If you don't understand the security implications of the
changes below then don't make them.

<OL>
<LI><STRONG><A NAME="item_Path_Configuration">Path Configuration</A></STRONG>
<P>
This is similar to the way <EM>/cgi-bin</EM> works - everything under this path is handled by SpeedyCGI. Add the
following lines near the top of your httpd.conf - this will cause all
scripts in your cgi-bin directory to be handled by SpeedyCGI when they are
accessed as <EM>/speedy/script-name</EM>.

<P>
<PRE>    Alias /speedy/ /home/httpd/cgi-bin/
    &lt;Location /speedy&gt;
        SetHandler speedycgi-script
        Options ExecCGI
        allow from all
    &lt;/Location&gt;
</PRE>
<LI><STRONG><A NAME="item_File_Extension_Configuration">File Extension Configuration</A></STRONG>
<P>
This will make SpeedyCGI handle all files with a certain extension, similar
to the way .cgi files work. Add the following lines near the top of your
httpd.conf file - this will set up the file extension ``.speedy'' to be
handled by SpeedyCGI.

<P>
<PRE>    AddHandler speedycgi-script .speedy
    &lt;Location /&gt;
        Options ExecCGI
    &lt;/Location&gt;
</PRE>
</OL>
<P>
<HR>
<H1><A NAME="FREQUENTLY_ASKED_QUESTIONS">FREQUENTLY ASKED QUESTIONS</A></H1>
<DL>
<DT><STRONG><A NAME="item_How">How does the speedy front end connect to the back end process?</A></STRONG><DD>
<P>
Via a Unix socket in <EM>/tmp</EM>. A queue is kept in a shared file in <EM>/tmp</EM>
that holds an entry for each process. In that queue are the pids of the
perl processes waiting for connections. The frontend pulls a process out of
this queue, connects to its socket, sends over the environment and argv,
and then uses this socket for stdin/stdout to the perl process.

</DL>
<DL>
<DT><STRONG><A NAME="item_If">If another request comes in while SpeedyCGI script is running, does the client
have to wait or is another process started?  Is there a way to set a limit
on how many processes get started?</A></STRONG><DD>
<P>
If another request comes while all the perl processes are busy, then
another perl process is started. Just like in regular perl there is
normally no limit on how many processes get started. But, the processes are
only started when the load is so high that they're necessary. If the load
goes down, the processes will die off due to inactivity, unless you disable
the timeout.

<P>
Starting in version 1.8.3 an option was added to limit the number of perl
backends running. See <STRONG>MaxBackends</STRONG> in <A HREF="#Options_Available">Options Available</A>
above.

</DL>
<DL>
<DT><STRONG>How much of perl's state is kept when speedy starts another request?
Do globals keep their values?  Are destructors run after the request?</STRONG><DD>
<P>
Globals keep their values. Nothing is destroyed after the request.
STDIN/STDOUT/STDERR are closed -- other files are not.  <CODE>%ENV</CODE> and <CODE>@ARGV</CODE>
are the only globals changed between requests.

</DL>
<DL>
<DT><STRONG>How can I make sure speedy restarts when I edit a perl library used
by the CGI?</STRONG><DD>
<P>
Do a touch on the main cgi file that is executed. The mtime on the main
file is checked each time the front-end runs.

</DL>
<DL>
<DT><STRONG><A NAME="item_Do">Do I need to be root to install and/or run SpeedyCGI?</A></STRONG><DD>
<P>
No, root is not required.

</DL>
<DL>
<DT><STRONG>How can I determine if my perl app needs to be changed to work with
speedy?  Or is there no modification necessary?</STRONG><DD>
<P>
You may have to make modifications.

<P>
Globals retain their values between runs, which can be good for keeping
persistent database handles for example, or bad if your code assumes
they're undefined.

<P>
Also, if you create global variables with ``my'', you shouldn't try to
reference those variables from within a subroutine - you should pass them
into the subroutine instead. Or better yet just declare global variables
with ``use vars'' instead of ``my'' to avoid the problem altogether.

<P>
Here's a good explanation of the problem - it's for mod_perl, but the same
thing applies to speedycgi:

<P>
<PRE>    <A HREF="http://perl.apache.org/docs/general/perl_reference/perl_reference.html#my___Scoped_Variable_in_Nested_Subroutines">http://perl.apache.org/docs/general/perl_reference/perl_reference.html#my___Scoped_Variable_in_Nested_Subroutines</A>
</PRE>
<P>
If all else fails you can disable persistence by setting MaxRuns to 1. The
only benefit of this over normal perl is that speedy will pre-compile your
script.

</DL>
<DL>
<DT><STRONG>How do I keep a persistent connection to a database?</STRONG><DD>
<P>
Since globals retain their values between runs, the best way to do this is
to store the connection in a global variable, then check on each run to see
if that variable is already defined.

<P>
For example, if your code has an ``open_db_connection'' subroutine that
returns a database connection handle, you can use the code below to keep a
persistent connection:

<P>
<PRE>    use vars qw($dbh);
    unless (defined($dbh)) {
        $dbh = &amp;open_db_connection;
    }
</PRE>
<P>
This code will store a persistent database connection handle in the global
variable ``$dbh'' and only initialize it the first time the code is run.
During subsequent runs, the existing connection is re-used.

<P>
You may also want to check the connection each time before using it, in
case it is not working for some reason. So, assuming you have a subroutine
named ``db_connection_ok'' that returns true if the db connection is
working, you can use code like this:

<P>
<PRE>    use vars qw($dbh);
    unless (defined($dbh) &amp;&amp; &amp;db_connection_ok($dbh)) {
        $dbh = &amp;open_db_connection;
    }
</PRE>
</DL>
<DL>
<DT><STRONG><A NAME="item_Why">Why do scripts with persistent Oracle database connections hang?</A></STRONG><DD>
<P>
When using an IPC connection to Oracle, an oracle process is fork'ed and
exec'ed and keeps the stdout connection open, so that the web server never
gets an EOF. To fix the problem, either switch to using a TCP connection to
the database, or add the following perl code somewhere before the
DBI-&gt;connect statement:

<P>
<PRE>    use Fcntl;
    fcntl(STDOUT, F_SETFD, FD_CLOEXEC);
</PRE>
<P>
This will set the close-on-exec flag on standard out so it is closed when
oracle is exec'ed.

</DL>
<P>
<HR>
<H1><A NAME="USING_GROUPS">USING GROUPS</A></H1>
<P>
The group feature in SpeedyCGI can be used to help reduce the amount of
memory used by the perl interpreters. By default groups are not used (group
name is ``none''), and each perl script is given its own set of perl
interpreters. Each perl interpreter is also a separate system process.

<P>
When grouping is used, perl interpreters and perl scripts are put in a
group. All perl interpreters in a group can run perl scripts in the same
group. What this means is that by putting all your scripts into one group,
there could be one perl interpreter running all the perl scripts on your
system. This can greatly reduce your memory needs when running lots of
different perl scripts.

<P>
SpeedyCGI group names are entities unto themselves. They are not associated
with Unix groups, or with the Group directive in Apache. Expect for the two
special group names ``none'' and ``default'', all group names are created
by the user of SpeedyCGI using the Group option described in <A HREF="#OPTIONS">OPTIONS</A>



<P>
If you want the maximum amount of grouping possible then you should run all
scripts with the group option set to ``default''. This the group name used
if you just specify ``-g'' on the command line without an explicit group
name. When you do this, you will get the fewest number of perl interpreters
possible - any perl interpreter will be able to run any of your perl
scripts.

<P>
Although using group ``default'' for all scripts results in the most
efficient use of resources, it's not always possible or desirable to do
this. You may want to use other group names for the following reasons:

<UL>
<LI><STRONG><A NAME="item_To">To isolate misbehaving scripts, or scripts that don't work in groups.</A></STRONG>
<P>
Some scripts won't work in groups. When perl scripts are grouped together
they are each given their own unique package name - they are not run out of
the ``main'' package as they normally would be. So, for example, a script
that explicitly uses ``main'' somewhere in its code to find its subroutines
or variables probably won't work in groups. In this case, it's probably
best to run such a script with group ``none'', so it's compiled and run out
of package main, and always given its own interpreter.

<P>
In other cases, scripts may make changes to included packages, etc, that
may break other scripts running in the same interpreter. In this case such
scripts can be given their own group name (like ``pariah'') to keep them
away from scripts they are incompatible with. The rest of your scripts can
then run out of group ``default''. This will ensure that the ``pariah''
scripts won't run within the same interpreter as the other scripts.

</UL>
<UL>
<LI><STRONG><A NAME="item_To">To pass different perl or SpeedyCGI parameters to different scripts.</A></STRONG>
<P>
You may want to use separate groups to create different policies for
different scripts.

<P>
For example, you may have an email application that contains ten perl
scripts, and since the common perl code used in this application has a bad
memory leak, you want to use a MaxRuns setting of 5 for all of these
scripts. You want to run all your other scripts with a normal MaxRuns
setting. To accomplish this you can edit the ten email application scripts,
and at the top use the line:

<P>
<PRE>    #!/usr/bin/speedy -- -gmail -r5
</PRE>
<P>
In the rest of your perl scripts you can use:

<P>
<PRE>    #!/usr/bin/speedy -- -g
</PRE>
<P>
What this will do is put the ten email scripts into a group of their own
(named ``mail'') and give them all the default MaxRuns value of 5. All
other scripts will be put into the group named ``default'', and this group
will have a normal MaxRuns setting.

</UL>
<P>
<HR>
<H1><A NAME="DOWNLOADING">DOWNLOADING</A></H1>
<P>
<HR>
<H2><A NAME="Binaries">Binaries</A></H2>
<P>
Binaries for many OSes can be found at:

<P>
<PRE>    <A HREF="http://daemoninc.com/SpeedyCGI/download.html">http://daemoninc.com/SpeedyCGI/download.html</A>
</PRE>
<P>
<HR>
<H2><A NAME="Source_Code">Source Code</A></H2>
<P>
The standard source code distribution can be retrieved from any CPAN mirror
or from:

<P>
<PRE>    <A HREF="http://daemoninc.com/SpeedyCGI/download.html">http://daemoninc.com/SpeedyCGI/download.html</A>
    <A HREF="http://www.cpan.org/modules/by-authors/id/H/HO/HORROCKS/">http://www.cpan.org/modules/by-authors/id/H/HO/HORROCKS/</A>
</PRE>
<P>
The latest development code can be obtained from the SourceForge CVS
repository using the following commands:

<P>
<PRE> cvs -d:pserver:anonymous@cvs.SpeedyCGI.sourceforge.net:/cvsroot/speedycgi login 
 cvs -z3 -d:pserver:anonymous@cvs.SpeedyCGI.sourceforge.net:/cvsroot/speedycgi co 2.x
</PRE>
<P>
Press Enter when prompted for a password.

<P>
<HR>
<H1><A NAME="AUTHOR">AUTHOR</A></H1>
<P>
<PRE>    Sam Horrocks
    <A HREF="http://daemoninc.com">http://daemoninc.com</A>
    sam@daemoninc.com
</PRE>
<P>
<HR>
<H2><A NAME="Contributors">Contributors</A></H2>
<P>
A lot of people have helped out with code, patches, ideas, resources, etc.
I'm sure I'm missing someone here - if so, please drop me an email.

<UL>
<LI>
<P>
Gunther Birznieks

<LI>
<P>
Diana Eichert

<LI>
<P>
Takanori Kawai

<LI>
<P>
Robert Klep

<LI>
<P>
Marc Lehmann

<LI>
<P>
James McGregor

<LI>
<P>
Josh Rabinowitz

<LI>
<P>
Dave Parker

<LI>
<P>
Craig Sanders

<LI>
<P>
Joseph Wang

</UL>
<P>
<HR>
<H1><A NAME="SEE_ALSO">SEE ALSO</A></H1>
<P>
<CODE>perl(1),</CODE> <CODE>httpd(8),</CODE> <CODE>apxs(8).</CODE>

<P>
<HR>
<H1><A NAME="MORE_INFORMATION">MORE INFORMATION</A></H1>
<P>
<HR>
<H2><A NAME="SpeedyCGI_Home_Page">SpeedyCGI Home Page</A></H2>
<P>
<A
HREF="http://daemoninc.com/SpeedyCGI/">http://daemoninc.com/SpeedyCGI/</A>

<P>
<HR>
<H2><A NAME="Mailing_List">Mailing List</A></H2>
<UL>
<LI>
<P>
SpeedyCGI users mailing list - <A
HREF="mailto:speedycgi-users@lists.sourceforge.net.">speedycgi-users@lists.sourceforge.net.</A>
Archives and subscription information are at <A
HREF="http://lists.sourceforge.net/lists/listinfo/speedycgi-users">http://lists.sourceforge.net/lists/listinfo/speedycgi-users</A>


</UL>
<UL>
<LI>
<P>
SpeedyCGI announcements mailing list - <A
HREF="mailto:speedycgi-announce@lists.sourceforge.net.">speedycgi-announce@lists.sourceforge.net.</A>
Archives and subscription information are at <A
HREF="http://lists.sourceforge.net/lists/listinfo/speedycgi-announce">http://lists.sourceforge.net/lists/listinfo/speedycgi-announce</A>


</UL>
<P>
<HR>
<H2><A NAME="Bugs_and_Todo_List">Bugs and Todo List</A></H2>
<P>
Please report any bugs or requests for changes to the mailing list.

<P>
The current bugs / todo list can be found at <A
HREF="http://www.sourceforge.net/projects/speedycgi/.">http://www.sourceforge.net/projects/speedycgi/.</A>
Go to the Bug Tracking menu and select the group ``bug'' for bugs, or the
group ``rfe'' for the todo list.

<P>
<HR>
<H2><A NAME="Japanese_Translation">Japanese Translation</A></H2>
<P>
<A
HREF="http://member.nifty.ne.jp/hippo2000/perltips/CGI/SpeedyCGI.htm">http://member.nifty.ne.jp/hippo2000/perltips/CGI/SpeedyCGI.htm</A>


<P>
<HR>
<H2><A NAME="Benchmarks">Benchmarks</A></H2>
<P>
<A
HREF="http://daemoninc.com/SpeedyCGI/benchmarks/">http://daemoninc.com/SpeedyCGI/benchmarks/</A>


<P>
<HR>
<H2><A NAME="Success_Stories">Success Stories</A></H2>
<P>
<A
HREF="http://daemoninc.com/SpeedyCGI/success_stories/">http://daemoninc.com/SpeedyCGI/success_stories/</A>


<P>
<HR>
<H2><A NAME="Revision_History">Revision History</A></H2>
<P>
<A
HREF="http://daemoninc.com/SpeedyCGI/CGI-SpeedyCGI/Changes">http://daemoninc.com/SpeedyCGI/CGI-SpeedyCGI/Changes</A>


<P>
<HR>
<H2><A NAME="YAPC_2001_Presentation">YAPC 2001 Presentation</A></H2>
<P>
I gave an Introduction to SpeedyCGI talk at YAPC 2001. It can be found at
<A
HREF="http://daemoninc.com/SpeedyCGI/yapc_2001/">http://daemoninc.com/SpeedyCGI/yapc_2001/</A>


<P>
<HR>
<H1><A NAME="COPYRIGHT">COPYRIGHT</A></H1>
<P>
Copyright (C) 2003 Sam Horrocks

<P>
This program is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the Free
Software Foundation; either version 2 of the License, or (at your option)
any later version.

<P>
This program is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
more details.

<P>
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc., 59
Temple Place - Suite 330, Boston, MA 02111-1307, USA.

<P>
This product includes software developed by the Apache Software Foundation
(http://www.apache.org/).

</BODY>

</HTML>