Predictably Scalability

Unique requirements make it more difficult to scale performance at the access edge than in the core of the network or at a network-to-network peering boundary. For example, the access edge system has to process registration traffic, secure subscriber connections, and protect the internal network from intrusions and attacks . Furthermore, as networks transition from connectionless (UDP-based) to connection-oriented (TCP-based) services, subscriber edge systems will be called upon to terminate increasing numbers of stateful connections.

Covergence Session Manager (CSM) is designed to predictably scale performance and capacity

High-performance multi-processor platforms – CSM mid-range and high-end systems use a high-performance multi-processor hardware platform equipped with a large amount of high-speed dynamic random-access memory (DRAM). A single CSM appliance can support hundreds of thousands of concurrent connectionless (UDP) or connection-oriented (TCP, TLS) subscriber sessions with high re-registration rates.

Parallel stream-processing architecture – CSM implements a patent-pending software architecture in which time-critical functions such as media processing are carried out at the operating system level, while less sensitive functions are executed at the application level. This parallelism lets CSM handle heavy signaling loads without impacting media performance.

Hardware-based media processing – Custom-designed MX-1 option cards execute media processing tasks on dedicated, specialized hardware. When equipped with two MX-1 cards, a single CSM appliance can handle up to 20,000 concurrent calls with negligible latency, jitter and loss.

High-availability clustering – Multiple CSM appliances can be organized into a high-availability, self-healing cluster. Clustering lets service providers scale CSM capacity to carrier-class levels without increasing management complexity. Automatic distribution of signaling and media streams across cluster elements ensures consistent performance as the offered load increases. State synchronization, network interface redundancy and shared service redundancy guarantee service continuity through equipment failures.



Distributed signaling and media processing – Other solutions back haul media flows from thousands of subscribers to a single server. But the back haul costs can be burdensome, and the central server becomes both a bottleneck and a single point of failure. With CSM, signaling and media processing functions can be integrated on a single appliance or distributed to separate appliances. Signaling and media appliances can be collocated or geographically dispersed, allowing optimal distribution of processing power for any situation.