FlowEngine™ for Datacenter Operators

Improve network service provisioning, economics, and flexibility

Cloud networking and virtualization are influencing the development and deployment of SDN and NFV solutions for carrier data centers. We designed our FlowEngine software technology to classify millions of packet flows entering the modern data center, and then intelligently distributing these flows to Physical or Virtualized Network Functions – all under SDN control. FlowEngine delivers programmable packet processing capabilities, including wirespeed DPI for rapid classification, integrated load balancing, service function chaining, flow affinity, SDN orchestration optimizations and more. 

FlowEngine helps Datacenter Operators solve the following challenges:

Challenge #1: SDN Scalability and Performance

SDN decouples control layer functions from underlying switching elements, delivering improved network agility, flexibility and centralized control. However, scaling SDN to support the switching of millions of data flows with centralized control is challenging. FlowEngine delivers an industry-leading 1.2 Tbps of throughput and a high rate of flow rule modifications in match/action tables to keep pace with application demands. FlowEngine also allows SDN controllers to anticipate and predefine flow processing logic, so that when new flows arrive, they can be autonomously forwarded against pre-defined logic, reducing latency for new flow processing by up to 50x. 

Challenge #2: Flow Classification for Network Analytics

Modern communication service providers are increasingly capturing network statistics and information in Big Data architectures for network analytics and forensices. FlowEngine can be configured by SDN controllers, using an OpenFlow interface, to identify, capture and forward packet flows according to intercept rules in the match tables. Once captured, FlowEngine can be programmed to pre-process packets, before forwarding them on to centralized Big Data storage.

Challenge #3: Realizing the Promise of NFV

Datacenter operators are leveraging NFV to implement virtualized network processing functions that can be turned up and down dynamically over cost-efficient virtualization and COTS server technologies. However, legacy systems are not designed to cost-effectively scale in an NFV environment.    

We designed FlowEngine technology to meet the scalabililty challenges of NFV deployments for both data center and telecommunications environments that demand high network throughput, scalability and guaranteed performance with flexible SDN control, all in a compact form factor. It enables CSPs to scale their infrastructure in real-time, accelerate on-demand service introduction and reduce CapEx and OpEx.

Why Datacenter Operators Choose FlowEngine™

Business Benefits

  • End-to-end application delivery
  • Quickly spin additional services to drive new revenue streams
  • “Filter and Tap” the required data at wire rates, allowing CSPs to provide “on-tap” services to their customers
  • Reduce CapEx and OpEx
  • Improve user experience with end-to-end automation of flow provisioning

Technical Benefits

  • Delivers Network Analytics with carrier-grade scale and performance
  • Fully programmable data and control plane
  • Integrated system with a rich software stack and elements that can be fully programmed to scale in/out on-demand with a cloud driven architecture

Common Questions from Datacenter Operators::

Yes. The vision of SDN orchestration is that each new flow gets reported to the orchestration layer for routing information. The reality is that this approach is challenging to scale, due to high probability of packet loss waiting for an SDN orchestration layer response, while overwhelming the orchestration layers.

FlowEngine blends SDN orchestration principles and architecture with the realities of high-speed data plane packet forwarding. Each flow is autonomously assigned to a VNF pool at wirespeed rates. FlowEngine also supports an in-band option for reassignment via orchestration platform.

Get In Touch