CS183 Lab 7: LAMP and Web Caching

Details: LAMP stands for Linux, Apache, MySQL and PHP, which is a very common combination platform for Web Programming. Squid Cache Server, on the other hand, is a very popular, high-performance and free web cache server.

You will need to use 2 VMs for this lab: the Router VM as a server and the Host VM as a client. You will need to install the Apache Web Server, MySQL, PHP and Squid Web Cache on the Router VM. Then you will verify them by using either wget or telnet from the Host VM.

Apache Web Server

• Install an Apache Web Server on the Router VM. Guess what the package name might be? Hint: do "yum list" to find out the name.
• Edit the config file /etc/httpd/conf/httpd.conf. Usually you would need to change "Listen" in the config file if the server will not be running on the standard HTTP port, ie. 80. Change "ServerAdmin" to your email address. Change "DocumentRoot" to whatever directory that stores your homepages.
• Before starting your web server, you need to disable your firewall temporary to allow HTTP connections from the Host VM.

  /etc/init.d/iptables stop
  

• Start the web server (see my updated www lecture slides).
• Create an /var/www/html/index.html on the Router VM, verify that your web server is working properly from the Host VM by using telnet and examine the access_log on the Router VM. Show the output in your report.
• Go to /var/www/html/, make a symbolic link to /etc/passwd
• Use telnet to "GET /passwd HTTP/1.0"
• Remove option "FollowSymLinks" from httpd.conf
• Reload httpd.conf
• Try to "GET /passwd HTTP/1.0" again, make sure it fails. Show the output in your report.
• Use chkconfig command to add httpd service into the startup scripts (see my updated www lecture slides). Reboot the Router VM to make sure the Apache Web Server starts automatically after rebooting. Show the output in your report.

PHP

• Install PHP on the Router VM. Make a guess again on what the package name might be!
• Make sure libphp5.so is installed under /usr/lib/httpd/modules/. Show the output in your report.
• Make sure php.conf is installed under /etc/httpd/conf.d. Show the output in your report.
• Restart your Apache web server (without rebooting).
• Create /var/www/html/index.php with the content below:

  <? phpinfo(); ?>
  

  Verify that your web server is able to serve PHP pages. Show the output in your report.

MySQL

• Install MySQL on the Router VM. You would need to yum install mysql-server and php-mysql.
• Create a text file named mysql.txt which is a script that creates a database (cs183) and a table (grades), and then inserts some data into the table. The content is as below:

    # Create a database named cs183
    create database cs183;
    # Select the database
    use cs183;
    # Create a grades table
    create table grades (name varchar(20), grade integer);
    # Add data into the table
    insert into grades values 
  	('Alice', 80),
  	('Bob', 90),
  	('Claire', 92),
  	('David', 60);
    

  and then run:

    $ mysql < mysql.txt
  

  to actually create the database and the table populated with data.

  NOTE: For security purposes, you would normally setup a password for the MySQL root account before using MySQL. However, we do not do that here. Please see here for details.

• Now, create a PHP script, /var/www/html/gradebook.php, that connects to the MySQL database and retrieves the information. The content is as below:

    <h1>CS183 Gradebook</h1>
    <pre>
    <?
  
    // Connect to the MySQL db
    $conn = mysql_connect("localhost", "root") or die(mysql_error());
  
    // Select the cs183 database
    mysql_select_db("cs183") or die(mysql_error());
  
    // SQL query
    $result = mysql_query("select * from grades order by grade desc;") or
  	die(mysql_error());
  
    // Display the result
    while ($row = mysql_fetch_assoc($result)) {
  	printf("%10s: %d\n", $row["name"], $row["grade"]);
    }
  
    // Free up resources
    mysql_free_result($result);
    mysql_close($conn);
  
    ?>
    </pre>
  

• From the Host VM, use either telnet or wget command to connect to the Router VM's web server and ask for the gradebook.php page. Show the result in your report and argue that your PHP+MySQL works.

Squid Cache Server

• Install Squid Cache Server on the Router VM. Download Squid from here.
• Login as a regular user (you might have to undo Lab 5 if nofile was set too low), un-bzip2, un-tar and compile the Squid:

  $ bzip2 -cd squid-2.6.STABLE20.tar.bz2 | tar xvf -
  $ cd squid-2.6.STABLE20
  $ CFLAGS="-O3" ./configure --prefix=/import/pkgs/squid
  $ make -j2
  $ make install
  $ pkginstall squid
  

• Edit /import/etc/squid.conf.
  • Change "maximum_object_size" to 60 MB.
  • Change "cache_mgr" to your email address.
  • Change "visible_hostname" to "cache".
  • Add the following lines after "our_netwworks" line:

    acl our_networks src 10.0.0.0/24
    http_access allow our_networks
    

• Create a directory /import/pkgs/squid/var/cache and change both the ownership and group ownership of it to "nobody".
• Change both the ownership and group ownership of /import/pkgs/squid/var/logs to "nobody".
• Run "squid -z" to create the swap directories (run only once!).
• Run "RunCache".
• From your Host VM, verify that your Squid Cahce Server is operating by using the telnet command. Show the output in your report.
• Again, from your Host VM, use wget with 'http_proxy' environment variable set to "10.0.0.1:3128" to download the lastest Linux kernel (eg. http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.25.tar.bz2) twice. Use the time command to measure the time taken to download the first and second time. Show that in your report. Hint: The second time should take a lot less time than the first time.
• Show the corresponding log entries from the /import/pkgs/squid/var/logs/access.log on the Router VM and explain it. Hint: Look for TCP_HIT or TCP_MISS associated with the URL you just downloaded.
• Add RunCache to /etc/rc.d/rc.local and reboot to make sure your Squid Cache Server starts automatically after rebooting. Show the output in your report.

Questions

1. Should the "FollowSymLinks" option in the web server config file be disabled by default? Why and why not?
2. Explain what <? phpinfo(); ?> does in the index.php?
3. Explain the code in gradebook.php. Show the result that generated by gradebook.php from the Host VM.
4. Why does the second download via the Squid Cache take a lot less time than the first time? Show the result from both downloads.

Scoring

• Attendance - 10%
• Question 1 - 10%
• Question 2 - 10%
• Question 3 - 20%
• Question 4 - 20%
• The rest of the report - 30%

• A sample lab report can be downloaded from here.