It's (Already) Baaack: Kelihos Botnet Rebounds With New Variant
Botnet hunters debate whether Kelihos/Hlux operators can reclaim rescued bots
Less than one day after botnet hunters announced they had crippled the Kelihos.B/Hlux.B botnet, a new version of the tenacious botnet is now back up and running today.
Researchers at Seculert were the first to point out the Kelihos/Hlux botnet was in action: Aviv Raff, co-founder and CTO at Seculert, late yesterday confirmed that his firm had seen the botnet spreading via a Facebook worm despite the announcement yesterday by Kaspersky, CrowdStrike, Dell SecureWorks, and The Honeynet Project that they had knocked the botnet offline. Raff says there's still communication under way via its command-and-control (C&C) servers.
More Security Insights
- Forrester Study: The Total Economic Impact of VMware View
- Securing Executives and Highly Sensitive Documents of Corporations Globally
- Top Big Data Security Tips and Ultimate Protection for Enterprise Data
- Smarter Process: Five Ways to Make Your Day-to-Day Operations Better, Faster and More Measurable
"We still see infected Kelihos.B machines, even new ones, sending spam and communicating with the C&C server," Seculert's Raff says.
But researchers from Kaspersky Lab, CrowdStrike, Fortinet, and Unveillance contend that this is a new variant of Kelihos/Hlux, not the same botnet that was taken down over the past few days. That one, KelihosB/HluxB -- which was built for spamming, information-stealing, DDoSing, as well as for pilfering Bitcoins and electronic wallets -- was sunk when the team poisoned it with their own code in order to redirect some 110,000 bots to their sinkhole server and away from the operator's control. It was about three times as large as the first Hlux/Kelihos botnet, which was crippled last fall by a team led by Microsoft and that included Kaspersky.
Marco Preuss, head of global research and analysis in Germany for Kaspersky Lab, says the new botnet activity is coming from Hlux.C/Kelihos.C, and that the one that his firm and others took down remains offline. "We confirm that a new Hlux/Kelihos sample exists, but it has a different configuration, which means it's coming from a new Hlux botnet (Hlux C). The previous generation (Hlux B) is under control by the sinkhole server. It is not uncommon for new versions of botnets to appear that are operated by the same group," Preuss says.
CrowdStrike's Tillmann Werner says the Kelihos.B-infected machines in the sinkhole set up by CrowdStrike and the other research teams can no longer be used by the botnet operators, and has confirmed that the C&C infrastructure doesn't use the Kelihos.B protocol anymore. The researchers say they diverted some 110,000 infected machines to their sinkhole server.
The new Kelihos.C variant, which Werner says was released "shortly" after the sinkhole operation began, is "completely separate" and is spreading via social networks. It contains a tweaked message format for spreading commands to its peers, he says.
CrowdStrike isn't aware of any way for the new botnet to steal back its bots from the sinkhole, he says.
Bit Seculert maintains that this new botnet activity is still Kelihos.B: "Some might call this 'a new variant' or Kelihos.C. However, as the new infected machines are operated by the same group of criminals, which can also regain access to the sinkholed bots through the Facebook worm malware, we believe that it is better to still refer this botnet as Kelihos.B," according to a blog post by the company that says more than 70,000 Facebook users were infected with the worm.
Debates aside, Kelihos/Hlux's comeback was not unexpected by the botnet hunters. But it was certainly faster than most had predicted. Why the quick turnaround?
Kyle Yang, manager of AV engine development at Fortinet, says the botnet operators obviously were ready for a takedown this time around. According to Yang's findings, the botnet operators had in their possession two "job" servers from the first Kelihos/Hlux variant taken down by Microsoft. Job servers are controlled by C&C servers.
"Pretty much those guys were ready for a takedown. They were just waiting for it," Yang says. "The reason why [they did it] so quickly was they still had two [job] servers alive" from the first botnet, he says. Those were not servers crippled by this week's botnet takedown, however, he says.
Yang says he hasn't been able to figure out just how the malware authors regained control of those two servers from the first round. The attackers also since have put up three additional job servers in support of the C version of the botnet, he says.
Unveillance, meanwhile, was able to confirm that the new variant uses a different protocol. "We can also confirm that the live sample we've been observing is what they are calling .C ... The two older binaries use well-formed HTTP and .C does not," says Karim Hijazi, CEO at Unveillance.
"In reference to Seculert's claim that because the same botmasters may be operating this new version that it should be called .B instead of .C, we disagree. Due to the change in the communication protocol, it seems perfectly in keeping with standard AV naming convention to give this a new version name," he says. "AV naming convention does not take into account distinct botnets within a given malware type. For example, the Mariposa C2 samples are still named Palevo/Rimecud, not Mariposa."
Most everyone does agree, however, that the only way to really kill a botnet is to apprehend its operators.
Next Page: Legal implications of taking over a P2P botnet?