How Nimda Changed Computer Security

This week marks the first anniversary of the Nimda virus attack, an event that may have driven more corporate IT security changes during the past 12 months than the September 11 terrorist attacks did.

Nimda first surfaced on September 18 last year and was among the first major viruses to target both servers and client computers. It combined features from previous threats and propagated not just via e-mail attachments, but also through shared files on servers. It also exploited Web pages containing Java scripts.

"Nimda heightened awareness, unfortunately at a very high cost," says Kim Milford, information security manager at the University of Wisconsin-Madison. For example, the virus showed that "filtering at the e-mail gateway or on the desktop alone wasn't the Holy Grail that we security folks are always seeking," Milford says.

Spreading Quickly

Among other things, the worm was capable of modifying certain types of Web documents, providing hackers with administrative access to systems, and creating back doors on infected systems that future attackers could exploit. It spread much faster and was quicker to cause damage than any previous worm or virus. According to antivirus vendor Symantec, there are more than 35,000 Nimda-related attacks occurring every day on corporate networks.

Nimda demonstrated the need for multiple layers of security more than any previous threat, says James M. Rinkel, senior vice president of systems services at Nova Information Services, a credit card processing firm in Atlanta.

One result is that companies have been forced to focus not only on network and perimeter security, but also on application- and database-level security, which Nova had been doing even prior to Nimda, Rinkel says. "It's also become crucial to have a plan to try and quarantine a virus if it gets into your systems, to keep it from spreading throughout the corporation," he says.

Stricter Security

After Nimda struck, the University of Chicago got stricter about removing improperly secured machines from its network, says E. Larry Lidz, a senior network security officer at the school.

"Before Nimda, we would alert the administrator of the machines that they were vulnerable to a security hole, but unless we had evidence that a machine was actually compromised, we generally left it on the network," Lidz says.

He adds that the university has also implemented a new process aimed at helping systems administrators install patches as soon as possible after security vulnerabilities are discovered in software that's widely used on its network.

"Nimda attacked the core content and data of enterprises," says Diane Fraiman, a vice president at Sanctum, a security software vendor in Santa Clara, California. "It brought home the fact that security is not just about network-level security or about authentication and authorization."

Much of this focus has resulted in increased spending on application-level intrusion-detection and firewall technologies in the year since Nimda struck, says John Pescatore, an analyst at Gartner.

Nimda also underscored the need for companies to install all patches recommended by software vendors, says Marty Lindner, a team leader for incident handling at the CERT Coordination Center in Pittsburgh. The worm succeeded because it took advantage of several well-known holes in popular software products, he says.