Cloud Computing Solutions

Efficient Layer 2-4 Pre- and Post- Processing Enables Internet-Scale Virtualization 

Xelerated's X11 and HX network processors enable Internet-scale virtualization in cloud computing environments. The technology is used for efficient L2-L4 classification, session management, pre- and post-processing in application-layer platforms.

Xelerated’s solution consists of a set of subsystems for optimized performance and scalability. Ecosystem partners provide layer 5-7 service and content recognition engines, while the Xelerated technology provides programmable layer 2-4 processing tasks.

Functional split for optimized performance in application platforms.

The split architecture approach has several benefits. Each functional part can be optimized for their respective tasks. Network Processors from Xelerated are for instanace 30 times more energy efficient compared to general purpose processors.

The X11 and HX network processors can be customized to any type of session pre- and post- processing, enabling scalable and flexible virtualization of services and cost-effective integration of service processing in router or switching platforms.

Range of Target Systems

• Intrusion Prevention Systems (IPS)
• Application policy enforcement systems
• Service gateways
• Virtual application servers
• Encryption and security gateways
• Server load balancers
• Virtual Ethernet Port Adapter (VEPA) systems
• Virtual Network Interface Cards (VNIC)
• Secure Socket Layer (SSL) Virtual Private Networks (VPN)
• User Access Control (UAC) systems
• Flow monitoring platforms
• LI systems
• Probe generation and performance monitoring systems
• High-speed stateful firewalls

Virtualized Server Line Card Design Example

Board layout example for a chassis-based cloud computing processing platform.

Feature Highlights

Application layer processing through partner engines and processors

• Content search/regular expression for URL check, cookies or malfunctional code
• Application level security for high-speed encryption

Application layer pre-processing

• Co-processing for flow classification
• Large-scale flow table memory access
• Port-level packet policing

Application layer post-processing

• Header modification and extension
• Fully flexible per flow metering, counting and statistics

Switching and data plane functionality

• Server and co-processor load balancing
• Efficient datapath and subsystem switching
• TCP splicing
• NAT/PAT
• IGMP/MLD snooping
• Stacking configuration support for pizzabox designs
• Traffic shaping and quality of service assignment
• OAM offload
• Multicast
• IP, MPLS, VPLS, and Ethernet bridging
• MAC address learning
• PPPoE and DHCP relay/proxy for subscriber session management