Like in the case of all apache modules, you can either compile PHP as a static module or compile it as a dynamic module.  In the case of static module , you  can’t perform any modification for the module without recompiling the binary to which it is attached.  For eg.  you can’t add ssl support for the mod_php without re-compiling apache as a whole. And any failure in the compilation may cause downtime for the entire webserver also, including plain html support.

The advantage is that it provides a faster  performance, because the module is initialized  whenever the apache binary is started.

In the case of a dynamic mod_php installation, the necessary modifications or module additions can be done by recompiling the module alone. There is no need to recompile the Apache as the mod_php is not linked with the binary of apache.  But since the webserver loads the module on the fly, it needs to load, initialize and then execute the module.  So it can create some level of slowness while processing php pages.

How does mod_php work?

When PHP  is loaded into Apache as a module (using mod_php), each Apache process will contain an instance of mod_php or PHP interpreter also.   The interpreter comes with a bundle of libraries we enabled during compilation and each  process can make use of these libraries to process the requests. This means that the Apache process that just started  to load a simple HTML page  too will contain a PHP interpreter with all assigned libraries which inturn means resource consumption.

When the webserver gets an HTTP request. The  request header  contains the path to the requested document

e.g. access.log:    xx.xx.xx.xx – - [22/June/2010:21:14:53 -0700] “GET /info.php?=PHPE9568F35-D428-11d2-A769-00AA001ACF42 HTTP/1.1″ 200 2146 “http://domain.com/info.php” “Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.1.7) Gecko/20100106 Ubuntu/9.10 (karmic) Firefox/3.5.7″

1. The request will be redirected to the document root of the domain and then to the file “info.php” , if it fails then the corresponding error message will be given.

2. The info.php file is to be processed. It follows the following steps

Normally every httpd.conf file will have an entry like this

AddType application/x-httpd-php .php5 .php4 .php

It instructs the webserver that the files with extension .php ,.php4,.ph5 are of PHP mime type . The TypesConfig directive sets the location of the MIME types configuration file. This file controls what Internet media types are sent to the client for  given file extension(s).  Sending the correct media type to the client  is important so they know how to handle the content of the file.

root@new [/usr/local/apache/conf]# cat mime.types | grep x-httpd-php
application/x-httpd-php-source          phps
application/x-httpd-php          php php3 php4 php5 php6
root@new [/usr/local/apache/conf]#

Here we can see that the Mime type to be used for files with extension .php  is application/x-httpd-php , while the file with .phps is to be mapped to the php mime type application/x-httpd-php-source .

The webserver identifies that the requested file is of Mime type x-httpd-php.

To handle or process it , the apache has to load the corresponding module. Since it is a php type , the module mod_php will be loaded and it will execute the file.

Since apache is a HTTP server. It gets the HTTP requests and answers with the HTML code.  So the mod_php will execute the commands within php flag and creates the HTML page dynamically and send it back to the client – internet browser which sent HTTP request.

Security concerns / Implications

You can see that  every request or execution of a php file through web is initiated by the webserver. So the webserver acts as the parent of every php execution through web. It imposes a great security threat. Since apache is being executed as an apache  user, all process will be owned by that user. By default  it is “nobody” or “apache”.  Let me try to explain.

If your  web application performs some operations in the db, unless that database (eg: a flat text DB) has built-in access control, you will have to make the database accessible to the “nobody” user. This means a malicious script could access and modify the database, even without a username and password.  Such can be the case with various configuration files too.  Unless you protect these directories or applications with necessary authorization techniques like .htaccess, session control etc. There is a high possibility of attack through webapplication.

Another dangerous issue is of root escalation. If the webserver has a bug, by exploiting that bug, a malicious user can gain some root privileges or escalated to root. Its quite alarming situation as an escalated apache user can do any sort of  actions without any level of authentication.

Also it is difficult to identify the script which performs the malicious activity as all php scripts will be executed as “nobody”

Since PHP applications are executed as web server user, you need to give access and write permissions for the directories wherever the application  is supposed to be working. Sometimes you may be forced to give 777 permissions and it invites lot of attacks.

The files created by php applications will be owned by user “nobody” . So the user will not be able to delete the files unless it is done through another php application. Otherwise he needs to contact the server admin to get the same.

As a security measure, we may be forced to block mails from “nobody” users . But it can create mails generated from php applications being blocked in the server. Various php applications widely used for spamming . So some servers are configured to block mails from nobody users. This creates inconvenience to users.

Tweet This Post

PHP is similar to many other scripting languages like perl, python etc. But unlike perl and python what makes it stand apart is, its adaptability and power to be used as both command line and server side scripting.  I hear Yahoo’s mail runs on PHP.

I shall try to explain you the difference by executing the same file on different modes. Don’t expect too much from this post.

Command Line (CLI)

PHP Command Line Interface or PHP CLI  as the name implies,  is a way of using PHP in the system command line, like below.

# php -i | more
phpinfo()
PHP Version => 5.2.10-2ubuntu6.3

System => Linux den 2.6.31-15-generic #50-Ubuntu SMP Tue Nov 10 14:54:29 UTC 200
9 i686
Build Date => Nov 26 2009 14:40:20
Server API => Command Line Interface
Virtual Directory Support => disabled
Configuration File (php.ini) Path => /etc/php5/cli
Loaded Configuration File => /etc/php5/cli/php.ini

In other words when the SAPI ( Server API) is “Command Line Interface” as you see above php-cli acts as a connector between underlying  php binary and the command / script which invokes php from shell . This enables processing of the scripts which require php functions and provide the result after execution of the script.   The advantage is that it doesn’t require a browser or webserver  for the execution. It simply requires a PHP parser ie PHP_CLI. This type of usage is ideal for scripts regularly executed using cron (on *nix or Linux) or Task Scheduler (on Windows). These scripts can also be used for simple text processing tasks. (Okay even image processing tasks )

PHP CLI is available on all popular operating systems.  The first step we need to ensure is that the CLI SAPI is installed. For this locate php binary path and execute the command as given below

[george@server3]$ type php
php is /usr/local/bin/php
[george@server3]$ /usr/local/bin/php -v | grep cli
PHP 5.2.9 (cli) (built: Jul 10 2009 19:22:08)

The result ensures that the php-cli module is installed. Now let us  see, how to make it work in a shell. The first thing is to identify the exact path on which php  binary is present. I assume that its /usr/local/bin/php.

Now create the test file info.php

vi info.php

<?php
phpinfo();
?>

Execute the  file from the shell using the command line or shell. If you intent to use it in a cron specify the binary path (interpreter) as the first entry of file .

/usr/local/bin/php info.php

Then the out put will be  of the following  format

[geroge@server3]# php info.php | head
phpinfo()
PHP Version => 5.2.9

System => Linux server3.xxxxxx.xxx 2.6.18-128.7.1.el5 #1 SMP Mon  Aug 24 08:21:56 EDT 2009 x86_64
Build Date => Jul 10 2009 19:17:35
Configure Command =>  './configure'  '--enable-bcmath'  '--enable-calendar' '--enable-dbase' '--enable-ftp'  '--enable-gd-native-ttf' '--enable-libxml' '--enable-magic-quotes'  '--enable-mbstring' '--enable-pdo=shared' '--enable-soap'  '--enable-sockets' '--enable-zend-multibyte' '--enable-zip'  '--prefix=/usr/local' '--with-curl=/opt/curlssl/'  '--with-tidy=/opt/tidy/' '--with-ttf' '--with-xmlrpc'  '--with-xpm-dir=/usr' '--with-zlib' '--with-zlib-dir=/usr'  '--with-litespeed'
Server API => Command Line Interface
Virtual Directory Support => disabled
Configuration File (php.ini) Path => /usr/local/lib
Loaded Configuration File => /usr/local/lib/php.ini

You can see the Application Programming Interface (API) for this script by  checking the variable Server API. In the above test, it is shown as Command Line Interface. The value gets changed according to the mode of usage or execution.

As I mentioned above the php binary was  invoked  by the php-cli interpreter. Once php binary executed the query , the result was returned to the SAPI  ie php-cli, it then passed the results to the command line.

Why do we use it?

Its a quite handy option to create cron jobs. Some times we may need to perform some sort of updates or script execution periodically. If we set  a cron job for this using php, the php-cli option is quite useful.

Server-side scripting

PHP is extensively used for creating dynamic Web pages. You create pages with PHP and HTML. When a visitor opens the page, the server processes the PHP commands and then sends the results to the visitor’s browser along with the static HTML pages. PHP code is embedded into the HTML source document and interpreted by a web server with a PHP processor module, which generates the web page document.

So the minimum requirement for here are the following ones

1.  Web browser ( client side)

2.  Web server

3.   The PHP parser :  which connects the webserver with the underlying php installation.

Illustration :

Now lets us, take the same example. Here I am trying to access the page through a web browser. Let us see how the result looks like

Here , you can see the difference. In the first case when I executed the same file from shell  I got the output without any format. Now the same result is shown as an HTML page.  So let us see  how the conversion takes pace.

The webserver is  connected to the php with the help of “Server API” . This is called PHP parser  it is either CGI or Server Module like apache, litespeed etc. You can see  it is “CGI” here, but it can vary according the mod of installation and webserver it uses. Common values are FastCGI/CGI , Litespeed API, Apache etc

Once the HTTP  / Web server gets the request for the page with php , it calls PHP interpreter to generate HTML. Then this HTML is returned to the client – internet browser which sent the HTTP request.

I believe you might got some ideas how it works or differs . I shall try to give  the difference between CGI and a server module on another post

Server API (SAPI)

PHP is meant  to work on all platforms. So it is essential that it should work with different types of webservers as well.   Every version of  php comes with a set of Binaries to connect php with various webservers and it is known as SAPI.  During configuration of php, we  need to give corresponding options for the webserver to which the interpreter is to be enabled. Once the php is compiled and installed , the corresponding SAPI  will be activated. So every communication between the webserver and PHP will be interpreted  using the corresponding SAPI module enabled.

Apache may use “Apache Handler” while LiteSpeed webserver use “LiteSpeed API” as the SAPI.

Extensions

Being a powerful language which can perform various levels of operation,  PHP needs to utilize various functions and libraries installed on the server. For example GD libraries are required for a php script which performs image manipulations. While a shopping cart or such applications require db connectivity to be enabled.  Like in the case of SAPI a lot of such connectors are shipped along with PHP.  To extend the functionality of php, we enable the necessary libraries while configuring the application using “–with”  option.

You can create your own php extensions and php modules and compile them also.  Again, we will have a discussion about it later.

Tweet This Post

Here we are going to install apache using the source only. The choice of Operating System here is Linux (distro: Centos). The procedure we follow here will lead to a simple apache installation for dynamic loading of php. PHP will be installed as a module to apache. With a little bit of patience and time, all the necessary modules can be installed with apache. I will be giving a brief idea about the installation of the other modules later.
The basics of installation from the source involves mainly three simple steps (assuming you are lucky):

./configure
make
make install

./configure creates the MAKEFILE on the fly. We can provide the necessary options to configure. To list the available options in the configuring step use

./configure --help

–prefix=/path/… mentions where the executable and its files are installed. If not mentioned it takes the default values.
–enable-[Feature] will enable the specified Feature in apache as it is being built. We are only interested in the DSO capability and hence we enable it with: –enable-so

As a convention we always keep the source code tar ball inside a directory in /usr/src/, thus source installation begins in this directory.

Apache Compilation

Download the required source tar ball of the apache you would like to compile. Here I am installing httpd-2.0.63 from http://httpd.apache.org/download.cgi#apache20. I save it in the /usr/src/ folder.

cd /usr/src/
wget http://www.bizdirusa.com/mirrors/apache/httpd/httpd-2.0.63.tar.gz

This will result in the generation of the file httpd-2.0.63.tar.gz

tar -xzf httpd-2.0.63.tar.gz

This will result in the creation of the directory httpd-2.0.63. Next enter inside this directory and execute the ./configure command.

cd httpd-2.0.63
./configure --prefix=/usr/local/webserver --enable-so

We are installing apache inside /usr/local/webserver and enable DSO to run php as a module to apache. During this process we may get a lot of errors. We resolve these by manually installing the unresolved dependencies either by obtaining their rpms or by using yum. Normally the first dependency we will get to resolve are:

gcc
glibc
libxml and
their corresponding devel packages

In the days where there were no package management tools like yum, pirut, apt-get etc. The old rpms served the installation of these packages with some effort. The task of determining the required rpm package for the required architecture and resolving the other dependencies which arise due to the installation of this rpm may be a tedious task. Some sites which helped in obtaining the necessary rpm suited for our installation and its other dependancies are:

http://rpm.pbone.net/

http://www.rpmfind.net/linux/RPM/

http://ftp.freshrpms.net/

http://dries.ulyssis.org/rpm/packages.html

http://apt.sw.be/

http://rpms.famillecollet.com/ (Remi RPM Repository)

Once everything goes well (we do the ./configure step again to determine this), the make command is executed.

make

If errors are encountered in this stage, We resolve them by installing the unresolved dependencies (Same as the previous step) and then do:

make clean

After this we repeat the make command and then issue:

make install

This process installs the package finally within the system. Modify the init script ( /etc/rc.d/init.d/httpd or /etc/init.d/httpd they are symbolic links) Or sometimes you may even have to create one from the apache site.

The following is the content of one such init script I have used. The line beginning with apachectl/some/path/here and httpd=/some/path/here have to replaced with the appropriate line we have used in the –prefix portion of ./configure.

#!/bin/bash
#
# Startup script for the Apache Web Server
#
# chkconfig: - 85 15
# description: Apache is a World Wide Web server.  It is used to serve \
#              HTML files and CGI.
# processname: httpd
# processname: httpd
# pidfile: /usr/local/apache2/logs/httpd.pid
# config: /usr/local/apache2/conf/httpd.conf
# Source function library.
. /etc/rc.d/init.d/functions
if [ -f /etc/sysconfig/httpd ]; then
. /etc/sysconfig/httpd
fi
# This will prevent initlog from swallowing up a pass-phrase prompt if
# mod_ssl needs a pass-phrase from the user.
INITLOG_ARGS=""
# Path to the apachectl script, server binary, and short-form for messages.
apachectl=/usr/local/webserver/bin/apachectl
httpd=/usr/local/webserver/bin/httpd
pid=$httpd/logs/httpd.pid
prog=httpd
RETVAL=0
# The semantics of these two functions differ from the way apachectl does
# things -- attempting to start while running is a failure, and shutdown
# when not running is also a failure.  So we just do it the way init scripts
# are expected to behave here.
start() {
echo -n $"Starting $prog: "
daemon $httpd $OPTIONS
RETVAL=$?
echo
[ $RETVAL = 0 ] && touch /var/lock/subsys/httpd
return $RETVAL
}
stop() {
echo -n $"Stopping $prog: "
killproc $httpd
RETVAL=$?
echo
[ $RETVAL = 0 ] && rm -f /var/lock/subsys/httpd $pid
}
reload() {
echo -n $"Reloading $prog: "
killproc $httpd -HUP
RETVAL=$?
echo
}
# See how we were called.
case "$1" in
start)
start
;;
stop)
stop
;;
status)
status $httpd
RETVAL=$?
;;
restart)
stop
start
;;
condrestart)
if [ -f $pid ] ; then
stop
start
fi
;;
reload)
reload
;;
graceful|help|configtest|fullstatus)
$apachectl $@
RETVAL=$?
;;
*)
echo $"Usage: $prog {start|stop|restart|condrestart|reload|status"
echo $"|fullstatus|graceful|help|configtest}"
exit 1
esac
exit $RETVAL

Next we add the daemon name to the list of services and run it. For this follow the below steps.

chkconfig --add httpd
chkconfig --level 2345 httpd on
chkconfig --list httpd
/etc/init.d/httpd start
lynx http://localhost/ will display the default apache page which means success.

PHP Compilation
Now we are going to install PHP 5.2.13 from source!
Go to /usr/src/

cd /usr/src/

Download the PHP source tarball and extract it

wget http://in3.php.net/get/php-5.2.13.tar.gz/from/in.php.net/mirror
tar -xzf php*
cd php*

Just like in the previous apache installation, we are going to do the ./configure step with the required setting which are displayed using

./configure --help

We are only interested in enabling php as a module (–with-apxs2) support for mysql (–with-mysql) and prefix line. So we go for:

./configure --with-apxs2=/usr/local/webserver/bin/apxs --with-mysql --prefix=/usr/local/webserver/php

The long command can be written in a shorter, more clearer format with:

./configure --with-apxs2=/usr/local/webserver/bin/apxs \
--with-mysql \
--prefix=/usr/local/webserver/php

The same instructions go for the errors here.
Once everything goes smooth:

make
make install

We can provide the recommeneded php.ini setting in the path /usr/local/webserver/php/lib (what ever is the –prefix + /lib) or just copy the recommended settings to /usr/local/webserver/php/lib (This file may have the name php.ini-recommended or php.ini-production)
cp php.ini-recommended /usr/local/webserver/php/lib/php.ini

From now on we can have php’s index page to be the default index page. For this in the apache’s config file append index.php to the directive – DirectoryIndex

The line would thus look like:

DirectoryIndex index.html index.html.var index.php

To make make apache call modular php to execute the php script when encountered, add the following lines to the conf file.

AddType application/x-httpd-php .php
DirectoryIndex index.html index.html.var index.php

Next to test your installation.
In the default document root, create a phpinfo file with the file name index.php an d the contents as:

<?
phpinfo();
?>

Now we will test the apache configuration for any syntax errors and then reload the apache webserver:

apachectl configtest (No errors should be reported)
/etc/init.d/httpd reload

Open a browser window and load the localhost as URL, we will be viewing the phpinfo page in here. In the phpinfo page, the portion Configure Command shows the actual compilation time options used while ./configure is used. The row corresponding to Server API mentions how the php is loaded. ‘Apache 2.0 Handler’ means that php was loaded as a module of apache. The rest of the values can be globally changed by making the required changes in php.ini or locally in .htaccess (which is possible only because it is loaded as an apache module).

Tips on installing PHP as a CGI

Here we do not require installing apache with the –enable-so option. A normal installation will do. The installation of php will not require the option –with-apxs2. However we will have to mention the location of apache source directory with –with-apache=../apache_1.3.14

In the httpd.conf file you will require adding:

ScriptAlias /php/ [path where your php folder is located]
AddType application/x-httpd-php .php
AddType application/x-httpd-php .php3
AddType application/x-httpd-php .phtml
Action application/x-httpd-php /php/php5

Tweet This Post