Original URL: https://www.theregister.com/2007/08/17/storm_worm_attacks/

Universities warned of Storm Worm attacks

Covering fire

By Robert Lemos, SecurityFocus

Posted in Security, 17th August 2007 07:02 GMT

Colleges and universities have come under attack by Storm Worm botnets following attempts to detect infections through vulnerability scanning, a response centre for academic networks stated last week.

The Research and Education Networking Information Sharing and Analysis Centre (REN-ISAC) sent out the warning last Thursday following "numerous incidents" and advised school information technology managers to respond quickly to any infection on their networks.

The Storm Worm's distributed denial-of-service (DDoS) attacks appear to strike back at the network of any computer that scans an infected system, REN-ISAC officials said in the advisory.

"The attacks have been ICMP (Internet Message Control Protocol), can last more than a day, involve a large number of sources scattered globally, and can yield very significant attack traffic," the advisory warned. "With the impending return of students for fall classes, the DDoS-the-scanner-when-scanned behavior represents a significant risk for the EDU sector."

The advisory is the latest warning regarding the Storm Worm, which first started spreading in January using fairly large, but controlled, bursts of email routed through previously compromised computers.

Each burst typically sends out a custom variant, trying to infect systems before the user updates their anti-virus definitions. The program compromises systems by luring their users into opening the attachments of messages with subject lines regarding news events, including violent storms in Europe - a characteristic that led to the program's naming.

The control exercised by the online criminal group seeding the program makes the Storm Worm's name a misnomer; the malicious software is not a worm, as it does not spread automatically, but a bot program. Because it relies on deception to infect target systems, security experts have also referred to the program as a Trojan horse.

Over the past seven months, the Storm Worm's methods have evolved. The program, which is responsible for sending a significant portion of the stock pump-and-dump spam seen on the internet, has used tensions in the Mideast as bait for its malicious hook and has varied the type of attachments used to hold its fraudulent advertisements.

The program has also adopted a technique known as fast-flux Domain Name Service (DNS) hosting to make it harder to take down the botnet.

The botnet created by the Storm Worm has also grown significantly. From January to the end of May, SecureWorks detected a total of 2,815 bots trying to send email to the company's clients. Between the beginning of June and the end of July, the cumulative number of internet addresses trying to send spam jumped to 1.7 million, averaging about 100,000 addresses each day.

While the Storm Worm has been used for denial-of-service attacks in the past, the ability to automatically attack any site that scans one of the botnet's nodes for vulnerabilities appears to be a recent development.

"I think they (the bot masters) are trying to prevent people from monitoring their bots and from downloading the (Storm Worm) file, because that is what they are trying to protect," said Joe Stewart, senior security researcher with SecureWorks.

The attacks are thought to be automated, rather than manual, because the surge in packets occurs almost immediately after an infected computer is scanned, he said.

"It is able to be triggered after a known set of actions," Stewart said. "It's not like the guy (bot master) wakes up three hours later and decides to DDoS in response."

The deluge of data comes after a vulnerability scanner attempts to check a compromised PC that is currently being used as a distribution node in the Storm Worm bot net. Such nodes serve up the files used by the Storm Worm to infect a PC and initiate a peer-to-peer connection with other infected systems. While the REN-ISAC warning specifically mentioned attacks on universities, Stewart would not say if companies had also been attacked.

Even school administrator balked at talking about the problem. Mark Bruhn, executive director for REN-ISAC, declined to describe the attacks or say who had been targeted, but said the warning needs to be heeded before information-technology departments become overwhelmed by returning students.

"We did pressure the parties involved in this situation to allow us to share quickly, because starting very soon, there will be millions of students returning to campuses with computers that have been connected to who-knows-where, and might be infected with who-knows-what," he said in a response to SecurityFocus' request for comment.

The REN-ISAC recommended that universities quickly respond to infections to minimise the chance that compromised systems become distribution nodes, which would trigger an attack. The group also recommended that schools monitor network utilisation for suspicious increases and have open communication channels with their internet service providers.

SecureWorks' Stewart also cautioned researchers to be careful of whom they scan.

"As a general rule, you always risk an attack whenever you go and probe people," he said. "Whether it is manual or automated, you always run that risk, so it is best to be prudent when scanning other networks, especially potentially malicious ones."

This article originally appeared in Security Focus.

Copyright © 2007, SecurityFocus