Chapter 17: HTTP: Hyper Text Transfer Protocol

Chapter 17

HTTP: Hyper Text Transfer Protocol

In this chapter:

	Examining HTTP server software on Apache
	Locating the different document trees for the system and user pages
	Finding more information

HTTP is the protocol used for Web surfing. Even if it is one of the younger branches of the TCP/IP protocol family, it's the one that made the Internet popular and useful for lay people. SuSE ships the world's most popular HTTP (Web) server, Apache, with its distribution. Apache is a derivative of the NCSA (National Center for Supercomputing Applications, University of Illinois, Urbana-Champaign) Web server, and started as a set of patches (A patch e) for this server. At a certain point, the providers of these patches formed a group of their own and1 released the first version, Apache 1.0, in December 1995.

You can find the history of Apache and an extensive set of documentation on the Web site of the Apache group at http://www.apache.org.

Apache has so many configuration options that even a brief introduction is beyond the focus of this book. SuSE's details on preconfiguring the server cover most of the standard features, and the server starts without incident with even a basic SuSE Linux installation. Apache uses five configuration files, located in /etc/httpd. The access.conf, httpd.conf, and srm.conf files are the main configuration files, holding all directives that control the server's behavior. There is no real reason to have three files. The directives are split up mainly for historical reasons and to separate different aspects of their duties into different files. The file magic holds data for the mod_mime_magic Apache module that is used to determine file types. Closely related to this issue is the definition of the MIME (Multipurpose Internet Mail Extensions) types in mime.types.

17.1 Walking through the document tree

The root directory of the document tree is set to /usr/local/http. SuSE provides a standard start page and links to online documentation as default documents. If you want to use your machine as a plain intranet or Internet server, this is the directory tree you have into which to put your documents, CGI scripts, or both. There is almost no need to change the configuration if you don't need any special features. The directory /usr/local/http has four subdirectories.

17.1.1 htdocs

This is where the html pages are stored. The file index.html represents the homepage of the server. You can create subdirectories, put more html pages into this directory, or both.

Everything in this tree will be accessible via HTTP if it's readable for the Web server. The server has the permissions of user wwwrun and group nogroup. So please make sure that all files you want to provide via the Web are readable by this user.

17.1.2 cgi-bin

Directory for CGI scripts. There are a few sample CGIs in this directory. Most useful is test.pl. It prints the server environment if you call it in a Web browser.

Try to load the page http://localhost/cgi-bin/test.pl. You will get something like this, which shows you all information about your server:

Date: Sun Feb 21 11:23:34 1999  
%ENV:  
SERVER_SOFTWARE = Apache/1.3.4 (Unix) SuSE/6.0 PHP/3.0.6 mod_perl/1.17  
DOCUMENT_ROOT = /usr/local/httpd/htdocs  
GATEWAY_INTERFACE = CGI-Perl/1.1  
UNIQUE_ID = NtBdNcCoAMQAAADoIj8  
REMOTE_ADDR = 192.168.0.66  
SERVER_PROTOCOL = HTTP/1.0  
SERVER_SIGNATURE = Apache/1.3.4 Server at Homer.suse.com Port 80  
REQUEST_METHOD = GET  
QUERY_STRING =  
HTTP_USER_AGENT = Mozilla/4.5 [en] (X11; I; Linux 2.2.1 i586)  
PATH = /bin:/usr/bin:/usr/ucb:/usr/bsd:/usr/local/bin  
HTTP_CONNECTION = Keep-Alive  
HTTP_ACCEPT = image/gif, image/x-xbitmap, image/jpeg, image/pjpeg,  
image/png, */* REMOTE_PORT = 4876  
HTTP_ACCEPT_LANGUAGE = en  
MOD_PERL = mod_perl/1.17  
SCRIPT_NAME = /cgi-bin/test.pl  
SCRIPT_FILENAME = /usr/local/httpd/cgi-bin/test.pl  
HTTP_ACCEPT_ENCODING = gzip  
SERVER_NAME = Homer.suse.com  
REQUEST_URI = /cgi-bin/test.pl  
HTTP_ACCEPT_CHARSET = iso-8859-1,*,utf-8  
SERVER_PORT = 80  
HTTP_HOST = Homer.suse.com  
SERVER_ADMIN = root@localhost

17.1.3 icons

The GIF files used by Apache to create directory listings are stored in this directory.

Unfortunately, the permissions for this directory are not correct in the standard installation. In order to make this work, you have to change them by setting the x flag for everybody:

# chmod a+x /usr/local/httpd/icons

17.1.4 include

Include files needed to compile additional Apache modules are stored here. You need them only if you want to extend the functionality of the server by adding your own modules or modules you got somewhere on the Internet.

17.2 Permitting user home pages

Besides having a main document tree, each user on the system can have his or her own homepage. The way SuSE sets it up makes this quite simple. All a user has to do is to create the directory public_html in his or her home directory. The file index.html in public_html represent the user's home page, and all relative links will be placed in this directory or a subdirectory of it. The URL to access a user's home is http://server-name/~user-name.

Apache has several possible setups for user home pages. You may want to change the setup if your Web server is on a machine other than that containing the user home directories. You could, for example, have a dedicated directory with subdirectories for each user. Please refer to Apache's documentation for more details.

17.3 More information

There are several good books about Apache that explain the overall concepts and give examples of different setups for different purposes:

	Apache: The Definitive Guide, edited by Ben Laurie, Peter Laurie, and Robert Denn
	Apache Server Bible, by Mohammed J. Kabir
	Apache Server for Dummies, by Ken A. L. Coar

Summary:

Apache is the HTTP server shipped with SuSE Linux. HTTP is the protocol used by the World103 Wide Web (WWW) for hypertext documents. SuSE's Apache setup needs almost no additional configuration. Its default setting is adequate for a standard WWW server. The server document tree is located in /usr/local/httpd/htdocs. In addition to these documents, all users on the system can have their own homepage by placing their documents into the subdirectory public_html of their home directory.