SonicWall TZ400 - Security appliance - GigE

Reassembly-Free Deep Packet Inspection (RFDPI) engine The RFDPI engine provides superior threat protection and application control without compromising performance. This patented engine inspects the traffic stream to detect threats at Layers 3-7. The RFDPI engine takes network streams through extensive and repeated normalization and decryption in order to neutralize advanced evasion techniques that seek to confuse detection engines and sneak malicious code into the network. Once a packet undergoes the necessary preprocessing, including SSL decryption, it is analyzed against a single proprietary memory representation of three signature databases: intrusion attacks, malware and applications. The connection state is then advanced to represent the position of the stream relative to these databases until it encounters a state of attack, or another "match" event, at which point a pre-set action is taken. As malware is identified, the SonicWall firewall terminates the connection before any compromise can be achieved and properly logs the event. However, the engine can also be configured for inspection only or, in the case of application detection, to provide Layer 7 bandwidth management services for the remainder of the application stream as soon as the application is identified. Extensible architecture for extreme scalability and performance The RFDPI engine is designed from the ground up with an emphasis on providing security scanning at a high performance level, to match both the inherently parallel and ever-growing nature of network traffic. When combined with multi-core processor systems, this parallel-centric software architecture scales up perfectly to address the demands of deep packet inspection at high traffic loads. The SonicWall TZ Series platform relies on processors that, unlike x86, are optimized for packet, crypto and network processing while retaining flexibility and programmability in the field - a weak point for ASICs systems.

This flexibility is essential when code and behavior updates are necessary to protect against attacks that require updated and more sophisticated detection techniques.