January 13, 2020

Mitigating TCP SYN floods

On the Nexusguard platform, you can configure protection from TCP SYN flood attacks. The mechanism works like this: When a client sends a connection request (SYN segment) to the host, the platform intercepts the SYN segment and responds to the client with a SYN/ACK segment. The platform waits the specified timeout period for the return ACK from the client to complete the TCP handshake. 


If the platform does not receive a return ACK during the timeout period, it drops the packet. If the platform receives a return ACK, indicating that the client is legitimate and is not spoofed, it establishes a connection with the requested server and forwards the original connection request. 


As a second layer of defence, the platform can be configured to limit the number of embryonic (half-open) connections. When the embryonic connection threshold of a connection is crossed, the platform acts as a proxy for the server and generates a SYN-ACK response to the client’s SYN request using the SYN cookie method. When the platform receives an ACK back from the client, it can then authenticate that the client is real and allow the connection to the backend server. 


Last but not the least, Nexusguard’s mitigation platform employs global BGP Anycast to disperse and mitigate attack traffic across the global scrubbing network, ensuring extreme resilience and low latency during attack time.

Get the latest cybersecurity news and expert insights direct to your inbox

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.