Together, thousands of computers can perform complex tasks that would crash a single machine.
When used for good, these linked machines are called a grid computer. When used for evil, they’re called a botnet.
Parabon Computation, a Virginia-based technology company, is using its grid computer to simulate the destructive effects of a botnet. This allows Parabon’s clients to analyze their networks’ abilities to withstand potential cyberattacks.
Such an attack occurred in August, when cyber-criminals paralyzed Twitter.com, rendering the social-networking website inaccessible for hours.
These criminals typically use spam emails and other tactics to spread malicious software to thousands — sometimes millions — of computers. The linked computers form a botnet, and the criminals can instruct them to log onto a specific website, overloading the site with traffic and shutting it down.
The tactic, called a denial-of-service attack, is used for political and monetary gain — or just to wreak havoc.
“After testing our clients’ systems, we’ll work with them to develop remediation plans,” says Steven Armentrout, Parabon’s founder and chief executive officer. “There are various challenges to guarding against denial-of-service attacks. You have to find IP addresses that are generating tons of traffic and divert them.”
Parabon leases idle space on unused computers at university laboratories, and the company links them into a grid that can process high-volume information. For instance, the Defense Department uses Parabon’s 10,000-computer grid to analyze its contract data, Armentrout says.
Parabon is now marketing its denial-of-service simulator, called Blitz, to government agencies and businesses that are likely targets for Internet criminals. With Blitz, Parabon’s thousands of linked computers log onto clients’ websites simultaneously. They overload the sites with traffic, and the clients then analyze their networks’ resiliency.
Armentrout demonstrated Blitz for National Defense. He instructed the program to hit one of his company’s websites 2,000 times per second. At first, the site loaded instantly. A minute later, the site wouldn’t refresh at all.