Home Page | Skip to Navigation | Skip to Content | Skip to Search | Skip to Footer

Why Your Perimeter Defense is Not Enough

Despite perimeter firewalls and anti-virus measures, worms and viruses are getting into internal networks with alarming success and doing tremendous damage. They spread fast, often swamping LANs within hours. Users are unable to do their work, and they overwhelm help desks with calls. The network and security administrators battling the outbreaks often spend weeks cleaning up systems and fully restoring operations. Once inside the network, worms can persist indefinitely.

So where are the worms coming from?
Certain perimeter defenses cannot recognize the threat. Worms and blended attacks often strike through the application layer of network communications. SoBig, Blaster, and Slammer are well-publicized examples. Half of the most critical Internet security vulnerabilities on the SANS Institute's 2004 top-20 list are application-layer vulnerabilities. Rudimentary perimeter firewalls, however, can only detect network-layer attacks.

Hackers work fast
The new breed of worms and viruses can propagate to every vulnerable host in a matter of minutes. New application and operating system vulnerabilities are discovered weekly, and hackers exploit them faster than administrators can choose the right patches to apply, test and deploy to both internal and remote computers. Because of these factors, it is a well-established fact that there will always be some PCs and hosts on the network that are in a vulnerable state. Patching and antivirus updating is impractical as a proactive security practice.

Even companies who are doing a great job strengthening perimeter defenses against the super worms are open to infection from the inside. Take for instance the Blaster worm which caused billions of dollars in damage due to lost productivity, system downtime, system recovery and continuing remediation after the attack. In many companies Blaster simply walked into the office with employees who plugged infected laptops into the corporate LAN and spread the infection to other devices on the network.

The pathways for internal infection are proliferating rapidly
The growing number of telecommuting and mobile employees connecting to the internal network using their own unmanaged PCs and other devices are blowing holes in traditional perimeter defenses. In addition, network guests-including contractors, business partners, and customer-are routinely given remote access to enterprises' Web-based applications and portals. While IT and security administrators have little or no control over the configuration of these endpoints, they are feeling the pressure to control endpoint security policy compliance.

Abuse by Insiders
Organizations desperately need effective solutions for ensuring the safety of all endpoints connecting to the LAN. They also need a solution for containing the internal spread of any worm that manages to sneak in. Additionally, there is one more aspect to internal security that demands attention. No internal security solution is complete that does not address controlling the access of insiders to applications and databases. Every enterprise has locks on file cabinets, yet on the LAN most employees have access to more information sources than they need for their jobs. At the same time, employees or contractors with ulterior motives have new hacking tools available to them that make it easy to penetrate business applications and the valuable data contained with them. And now attackers are exploiting application layer protocols, flaws in application business logic, and poor authentication controls-further compounding the problem.