DNSSEC Can Encourage DDoS – Nexusguard

Nexusguard is, it says, "a cloud-based distributed denial of service (DDoS) security solution provider."

It put out a "Threat Report Distributed Denial of Service (DDoS) for Q219" that is understandably focused on the part of the security problem that it routinely deals with. The company says it gets the numbers reflected in the report from "attack data, research, publicly available information, Honeypots, ISPs, and logs recording traffic between attackers and their targets." All this effort goes to evaluate DDoS events "in a manner that is not biased by any single set of customers or industries," as the report put it.

Nexusguard found that DNS Amplification contributed to the largest share, compared to other methods, of attack activities in Q2 2019. It showed up in 65.95% of the attacks.

A DNS Amplification attack means that UDP packets with spoofed target IP addresses are maliciously sent to a publicly accessible DNS server. Attempting to respond to the sender, DNS resolvers transmit a large response to the target's spoofed IP address. The target receives an enormous amount of responses from the surrounding network.

Consider for context that during the quarter, Nexusguard's honeypot network captured 144,465,553 malicious DNS queries. That's a lot, which is sort of the point of the amplified attack.

It's no wonder that Nexusguard found that DNS Amplification was the leading DDoS attack vector, showing sharp increases of 31.01% QOQ and 1,040.41% YOY, respectively. HTTP and HTTPs Flood followed, dropping 12.78% and 36.00% (QOQ), while increasing 281.51% and 363.33%, respectively, (YOY).

The HTTP Flood attack is the second most common attack, attempting to exhaust server resources by generating valid, volumetric HTTP requests or sessions. The most common method to launch these attacks is HTTP GET flooding.

The third most common vector is HTTPS Flood. These sessions are typically HTTPS GET. They overwhelm the victim's web servers by flooding them with answer requests (ACK). The process forces servers to allocate maximum resources to handle the volumetric attack traffic. Normal traffic can't get though because of this.

Nexusguard found single-vector attacks dominated with 63.56% of the total, while multi-vectors accounted for the rest. Two- and three-vectored attacks were 13.56% and 8.71%, respectively. The maximum number of vectors they saw was 13.

As far as duration of the attacks, 74.18% of attacks lasted fewer than 90 minutes. 2.42% lasted more than 1,200. The quarterly average was 182.9 minutes, while the longest attack they saw lasted 28 days, 1 hour, and 11 minutes. In the quarter, the average duration dropped by 65.57% (QOQ) and 42.50% (YOY) and the maximum duration fell by 3.76% (QOQ) while rising 467.97% (YOY).

The report has an interesting take on DNSSEC. It notes that, "The growing adoption of DNSSEC suggests that DNS Amplification will continue to pose a significant threat to service provider and enterprise networks alike. Long overdue, the deployment of DNSSEC as the ultimate patch for fixing DNS cache poisoning is finally gaining widespread acceptance. The downside is the exceptionally long responses DNSSEC-enabled servers generate. The long DNS responses include records containing cryptographic keys and/or signatures. When a domain is upgraded to support DNSSEC, it returns traditional records as well as DNS records. As a result, the sizes of DNSSEC-enabled DNS responses significantly exceed those of traditional responses."

An attacker will like this, since it is easier for an attacker to stuff up a system with a long response. But the rest of us may put up with the potential problem of DDoS for the cache fix of DNSSEC.

— Larry Loeb has written for many of the last century's major "dead tree" computer magazines, having been, among other things, a consulting editor for BYTE magazine and senior editor for the launch of WebWeek.