Abstract: A method for providing network services includes the ability to dynamically and flexibly configure resources in a network server platform to service classes and sets of network traffic. According to one aspect, a programmable network server platform can be configured to bind multiple traffic classes in a mixture of inline and tap modes to a group of CPUs. Traffic having the associated classes can be configured to pass directly through while sending a copy down to one or more CPUs, as well as load balanced to a group or sub group of CPUs. According to another aspect, different applications or configurations can be run on different groups or sub groups of CPUs allowing complete flexibility.

Abstract: A system as described herein offers flexible and high-performance packet processing, and a unique scaling technology that matches the resource requirements with the demands of the individual application and operating environment. According to one aspect, the architecture enables wire-speed execution of any Packet Handling application, at any packet size. According to another aspect, the invention provides a general-purpose computing platform that enables Packet Handling applications to be dynamically developed and put into service, for a configurable set of traffic and with configurable amounts of processing power. According to a further aspect, the dynamically scalable architecture of the invention enables the processing power of the appliance to be freely adjusted to the performance requirements of a given application, without requiring any special configuration changes.

Abstract: A method for providing network services includes the ability to dynamically and flexibly configure resources in a network server platform to service classes and sets of network traffic. According to one aspect, a programmable network server platform can be configured to bind multiple traffic classes in a mixture of inline and tap modes to a group of CPUs. Traffic having the associated classes can be configured to pass directly through while sending a copy down to one or more CPUs, as well as load balanced to a group or sub group of CPUs. According to another aspect, different applications or configurations can be run on different groups or sub groups of CPUs allowing complete flexibility.

Abstract: A system as described herein offers flexible and high-performance packet processing, and a unique scaling technology that matches the resource requirements with the demands of the individual application and operating environment. According to one aspect, the architecture enables wire-speed execution of any Packet Handling application, at any packet size. According to another aspect, the invention provides a general-purpose computing platform that enables Packet Handling applications to be dynamically developed and put into service, for a configurable set of traffic and with configurable amounts of processing power. According to a further aspect, the dynamically scalable architecture of the invention enables the processing power of the appliance to be freely adjusted to the performance requirements of a given application, without requiring any special configuration changes.

Abstract: A method for providing network services includes the ability to dynamically and flexibly configure resources in a network server platform to service classes and sets of network traffic. According to one aspect, a programmable network server platform can be configured to bind multiple traffic classes in a mixture of inline and tap modes to a group of CPUs. Traffic having the associated classes can be configured to pass directly through while sending a copy down to one or more CPUs, as well as load balanced to a group or sub group of CPUs. According to another aspect, different applications or configurations can be run on different groups or sub groups of CPUs allowing complete flexibility.

Abstract: Systems and methods are described for converting priority based rules into isomorphic longest match rules. Rules for packet processing may be presented to a networking device in priority order, through an interface such as a Command Line Interface (CLI) or from networking applications which may reside on the networking device. The networking device may include hardware and/or software layers for accelerating packet processing; a forwarding layer may include hardware and/or software designed to perform longest match searches on packets. Prioritized rules may be converted into a data structure for the forwarding layer, so that a longest match search performed by the forwarding layer on the data structure is equivalent to a priority order search on the prioritized rules.