IPv6 To Complicate Threat-Intelligence LandscapeReputation-based blacklists could face exponential growth when the number of possible Internet addresses becomes, for all practical purposes, infinite
A common type of Internet-based threat intelligence is the assigning of reputation scores to the source of traffic, usually expressed as a certain Internet address or domain.
Yet with the gradual -- some would say "glacial" -- move to the Internet Protocol Version 6 (IPv6) address scheme, the Internet's address space will grow from merely big to nearly infinite. The vastness of the address space will cause problems for many threat-intelligence firms, from allowing attackers to use a new address for every attack to causing a rapid expansion in the size of the database needed to track the data on various sources, says Tommy Stiansen, chief technology officer for Norse, a real-time threat intelligence provider.
"IPv6 makes the whole thing interesting because it's a lot bigger," Stiansen says. "Databases will have to be rearchitected to handle the increased data. For anyone in threat intelligence, that will be the biggest challenge to overcome."
A small, but still significant, part of the Internet has adopted IPv6. While the global rate of adoption is a mere 1.6 percent, according to statistics provided by Google, about 4 percent of networks in the United States have an end-to-end implementation of IPv6. Moreover, the fraction of networks that use IPv6 is growing exponentially.
While security researchers have already raised general concerns on the impact that the expansion of the Internet's address space will have on various security technologies and techniques, the impact on the services that collect data on attackers and their activities has not generally been studied. Because the Internet address of malicious traffic is often used as a proxy identifier for the attacker and as a way to turn intelligence into action -- blocking traffic from the offending IP address -- IPv4 addresses are often a key component of the first line of network defense.
[With IPv6, a deluge of new top-level domains, and DNSSEC all coming, the Internet will become a much bigger place, meaning that defenses that worked in the past won't work in the future. See Broader Digital Landscape Means More Places To Hide.]
For the current Internet addressing scheme, the model works fine, says Craig Sprosts, vice president of product management for domain-name system (DNS) firm Nominum.
"A lot of security technologies and intelligence vendors rely on an IPv4 address or range of addresses -- it tends to be a reasonably effective way of representing identity of the attacker," Sprosts says. "It is a very reliable indicator of where the attack is coming from for certain protocols."
Yet there will be no scarcity of IPv6 addresses, and so attacker may frequently change their assigned numbers, resulting in a potentially fast expansion of the IP address space that vendors deem to be suspicious. Threat intelligence systems that assign a reputation score to an Internet address will have to be rethought, says Brian Foster, chief technology officer for network-security firm Damballa. Managing the reputation of the 4.2 billion addresses available under IPv4 is possible, but doing it for the vast IPv6 address space is not, he says.
"With IPv6, that goes right out the window," Foster says. "A lot of the reputation systems that are simple blacklists and are signature-based are going to have to be rethought and redone."
Damballa has focused on creating algorithms that can generate reputation scores on the fly, lessening the need for large databases of scores.
The problems are not limited to the increase in data requirements needed to track malicious behavior. Attackers will also find ways to hide in IPv6 traffic that are not available to them in Internet protocol version 4. For example, using extension headers -- a feature of IPv6 designed to support additional functionality -- an attacker could include a payload disguised as a Web packet, for example, but, in reality, also includes attack code, says Jeremy Duncan, senior director and IPv6 network architect for Salient Federal Solutions.
"If I rewrite the code for malicious software to use the extension header, normal intrusion detection systems can't see those headers, and I've evaded defenses," he says.
Those problems, however, should eventually be resolved as companies figure out how to handle IPv6 traffic and make their products more IPv6 aware, Nominum's Sprosts says.
"Over time this risk will be mitigated, but I think over the next few years it will be a problem," he says.
Have a comment on this story? Please click "Add Your Comment" below. If you'd like to contact Dark Reading's editors directly, send us a message.
Robert Lemos is a veteran technology journalist of more than 16 years and a former research engineer, writing articles that have appeared in Business Week, CIO Magazine, CNET News.com, Computing Japan, CSO Magazine, Dark Reading, eWEEK, InfoWorld, MIT's Technology Review, ... View Full Bio