Web Server Exploits

Web servers host Web pages that are made available to others across the Internet or an intranet. Public Web servers (those accessible from the Internet) always pose an inherent security risk because they must be available to the Internet to do what they are supposed to do. Clients (Web browser software) must be able to send transmissions to the Web server for the purpose of requesting Web pages. However, allowing transmissions to come into the network to a Web server makes the system-and the entire network-vulnerable to attackers, unless measures are undertaken to isolate the Web server from the rest of the internal network.

Web server applications, like other software, can contain bugs that can be exploited. For example, in 2001 a flaw was discovered in Microsoft's IIS software that exploited the code used for the indexing feature.The component was installed by default. When it was running, hackers could create buffer overflows to take control of the Web server and change Web pages or attack the system to bring it down. Microsoft quickly released security patches to address the problem, but many companies do not upgrade their software regularly nor do they update it with available fixes as they become available. New and different security holes are being found all the time in all major Web server programs. For example, major flaws have also been found in Apache Web servers' Hypertext Preprocessor (PHP) scripting language that, if exploited by an attacker, can result in the attacker running arbitrary code on the system. Security patches are available to address these and other issues, but that doesn't mean they are actually applied to the system.

The issue with vulnerabilities is also common in the platforms on which Web servers run, making a Web server vulnerable at its very foundation. For example, in 2005, the Zotob Worm infected numerous systems (including those of CNN and the Department of Homeland Security) days after a patch had been released addressing the plug-and-play vulnerability it exploited. While it would be nice to think that these were exceptions to the rule, this often isn't the case. Many administrators are remiss in identifying security holes quickly and installing the necessary software to fix the problem. Even worse, they may have unpatched older systems that still contain vulnerabilities that are several years old, and ripe for a hacker to attack.Web server exploits are popular for numerous reasons. One such reason is because firewalls are usually configured to block most traffic that comes into an internal network from the Internet, but HTTP traffic usually is not blocked.There are a large number of HTTP exploits that can be used to access resources that are outside the webroot directory.These include the Unicode Directory Transversal Exploit and the Double Hex Encoding Exploit.These are used to "sneak" the "../" directory transversal strings past the server's security mechanisms, which generally block URLs that contain the string. Another reason these exploits are so popular is that it's not necessary for hackers to have sophisticated technical skills to exploit unprotected Web servers. Scripts to carry out buffer overflow attacks, for example, can be downloaded and executed by anyone.

These are just a few examples of the ways that Web servers can be exploited, making it vitally important that these machines be secured. In addition to best configuration practices, there are software packages that are designed specifically to protect Web servers from common attacks.