CSCI A348/A548 Lecture Notes 3 Fall 1999

Review: Basics of Web Communication

• the display of an HTML document involves an interplay between the web user, browser, and server. A CGI script complicate this picture only at the server end; the user and the browser may have no knowledge that they are causing a program to run

• a web client communicates with a web server using the hypertext transport protocol (HTTP). HTTP transactions perform simple actions like retrieving an HTML page or executing a script

• CGI, or the Common Gateway Interface from the browser's point of view the execution of a script is identical to the retrieval of a static page. All communication with the script is mediated by the server

• from the browser's point of view the execution of a script is identical to the retrieval of a static page. All communication with the script is mediated by the server

• scripts are typically distinguished from ordinary HTML for example by their location

• the output of each script typically starts with a Content-type header which identifies the Media Type of the script output. The header must be followed by a blank line.

A review of Apache install

We have described the basic directory structure and have identified three important elements:

• the web server program (an executable)
• the web server root directory
• the web server document root directory

We have identified the conf directory and three configuration files:

• httpd.conf, that contains information that the server needs at startup
• srm.conf that contains information that the server needs while it runs
• access.conf defines the access to the files on your site

• specified the server root and the port number in httpd.conf
• specified the locations of the script and document root directories (in srm.conf)
• specified two path names in access.conf for the document root and scripts directories.

We described how the server can be started and stopped.

We mentioned that we need to be able to restart the server automatically if we need to be on-line continuously, and we have said we will describe ways to do that using cron.

The process id (pid) of the http daemon is located in a file httpd.pid in the logs directory.

The same directory contains a file where accesses are logged and a file where errors are recorded. We looked at them and described them briefly.

Java and Perl programs

I will write the minimal, shortest programs in Perl and Java, compile and run them for you. I will also use the Perl script as a CGI application.

HTTP

We will use telnet to request files from the server.