Ready or Not, IPv6 Security Threats are Coming
by VPNHaus | 08/17/2010 | Industry Commentary
There’s a simple math problem causing quite a lot of pain for companies who use the Internet. Here’s the math: seven billion does not equal four billion. As simple as this statement is, the complexity it creates is staggering. IPv4 represents the smaller sum. The solution, of course, is IPv6 with its 128-bit scheme, compared to the 32-bit predecessor. That equates roughly to 3.4x1038 unique addresses, enough to cover the seven billion people on the planet today and more than enough to substantially future-proof the protocol until we’re all well done and gone.
The security threat for companies in this situation lies in how to update all the technology to reflect the inevitable shift to IPv6. This includes all the technology they rely on that runs, processes or navigates any Internet data stream.
First, let’s cover the baked-in security of IPv6 protocol stack. In simple terms, the major difference is section RFC4601 which mandates use of IPsec for all nodes – something available for IPv4, but not required. The large address space in IPv6 safeguards against port scanning. Again, there’s math here that Samuel Sotillo details in his East Carolina University paper. Changes to the authentication header; encapsulating security payload, transport and tunnel modes; protocol negotiation and key exchange; and neighbor discovery and address auto-configuration further improve security.
Defcon speaker, Sam Bowne warns the industry that IPv6 adoption will likely cause “severe security headaches” because IT professionals haven’t really dug into the issue yet, as it’s not widely adopted today. What is happening today is a slow rollout – or a dual-stack environment – where both v4 and v6 are co-mingling, creating two infrastructures to secure, instead of just one. Bowne stressed during his presentation that it is extremely important for white-hat hackers to dig in and identify these threats. Sotillo also identifies a few areas worthy of inspection, including header manipulation issues such as spoofing, and flooding issues such as Smurf-type attacks on multicast traffic. Jake Kouns and Daniel Minoli dive into these issues in detail with their 2008 book, Security in an IPv6 Environment.
Interestingly enough, much of the advice given as far back as 2005 has still not been widely adopted. For example, Mike Chapple, CISSP, offered five tips that networking pros should pay attention to, including education across configuration, new tunneling protocols risks, and addressing complexity created by auto-configurations. Yet most professionals are still unfamiliar, according to a recent article by Robert Westervelt of SearchSecurity.com.
Buffer overflows and bugs will be an issue with the IPv6 transition as well. Joe Klein, Defcon attendee and subject matter expert with the North American IPv6 Task Force, states that it will take years for the bugs and flaws to be worked out.. But these will do, as it starts to gain wide acceptance. One particular flaw that is unique to IPv6 and causes chaos in networks is packet amplification attacks. This particular attack places a "0" in the routing header of each packet, and causes them to travel in a looped path. Ping pong exploits then take advantage of the 64 subnets available in the protocol, and allows attackers to send packets from one non-existent connection to another. This results in an ongoing series of ICMP Unreachable error messages and floods the network with wasteful data. In a podcast with TechRepublic’s Michael Kassner, Klein gives a great overview to of some of other issues that’s worth a listen.
IPv6 is a completely new protocol, not a simple patch slapped on existing IPv4 technology. Any technology has to be able to handle these changes, including VPN, routers, intrusion detection and prevention, firewalls, and network access control (NAC) solutions. Work-around solutions create gaps and gaps are what hackers exploit.
[tweetmeme source=”vpnhaus” only_single=false]