Issue description

Working with traditional hardware-based networking elements requires the procurement then physical installation and wiring of the hardware, which could take weeks and does cost very real money in labor and truck-rolls.

Virtualizing networking functions greatly simplifies this process and shrinks the availability ramp-up time of a new networking function from weeks to literally minutes. Another advantage of a VNF vs traditional hardware-based NEs is that it can be re-cycled instantly if its functions are no longer necessary.

However, from the provisioning and management point of view, working with virtualized networking functions is more complex than working with a dedicated NE. In addition to the hardware layer, VNFs require a hypervisor level to function. And on top of its underlying OS, it has to have the necessary networking functions to efficiently support the VNF.

Solution

Netlayer’s NFVgrid provides the necessary set of functions to instantiate, provision, monitor, troubleshoot, add/remove resources, and re-cycle Virtual Networking Functions. It manages the entire life-cycle of the VNF from spawning to dismantling.

By supporting the most popular hypervisors, NFVgrid can run the vast majority of VNFs available on the market. In addition, NFVgrid supports cloud platforms such as OpenStack, thus allowing VNFs to efficiently function in multi-location and software-based networking environments.

In contrast with hardware-based NEs, VNFs can be scaled up and down to achieve the highest resources utilization without impacting the functions it provides. This elasticity allows just-in-time reactions on hardware resources given bandwidth spikes and steady increases/decreases.

NFVgrid provides fully automated functions to scale a VNF up or down depending on its viable resource utilization: CPU usage, Memory utilization and Networking throughput.

Functions

• VNF instantiation and initial provisioning. VNF instantiation from stored images, with initial provisioning done by NFVgrid according to the provisioning request. In addition, NFVgrid registers the VNF with the hypervisor/cloud networking and/or SDN and configures them appropriately;
• VNF provisioning and initialization. Post-instantiation provisioning to connect with underlying networking resources;
• Monitoring. NFVgrid constantly monitors the VNFs under its management as well as server/software resources where these VNFs are running. Monitoring data is aggregated and available for users via UI or API;
• Troubleshooting. NFVgrid provides a rich set of troubleshooting methods including Cut Through functions and flow analytics;
• On-the-fly scaling. NFVgrid’s provisioning engine is based on the information received from the performance management system.

Scenarios

• Networking functions ready to go in minutes;
• Customers (internal or external) can add, change and re-cycle resources instantly;
• Networking resources are scaled up and down automatically; no need to undersubscribe.