Breaking Down DDoS UDP Flood Attacks
A UDP flood is a type of volume based DDoS (Distributed Denial of Service) attack, where large numbers of UDP (User Datagram Protocol) packets are sent to a target server, limiting its ability to carry out its functions. UDP traffic doesn’t require a three-way handshake to make a connection like TCP (Transmission Control Protocol), it runs with lower overhead and is ideal to carry data that doesn’t need to be checked and rechecked, such as VoIP. This means it is easier for attackers to generate large traffic volumes with tools like Low Orbit Ion Cannon (LOIC) and UDP Unicorn.
When using this type of DDoS attack the primary aim is to overwhelm the target network with packets to random UDP ports with a forged source IP address. These requests force the target host to look for the application that is running on those random ports (which may or may not exist) and flood the network with Internet Control Message Protocol (ICMP) destination unreachable packets, thereby blocking legitimate requests.
This attack can be managed by deploying perimeter defences such as Intrusion Detection Systems in a network to filter out unwanted network traffic. The target network would then never receive and never respond to the malicious UDP packets because the IDS would stop them. However, as there is a limit to the number of sessions managed by most IDS’s, they can also be susceptible to a UDP flood attack.
It is thought over 56% of DDoS attacks are UDP floods, which is why DDoS attack visibility is so important. Security analysts need to right tools to enable them to quickly determine the origins of an attack, trace it’s footprint in the network, identify the type of attack vector and if it is masking something more sinister such as a data exfiltration.