Network functions virtualization (NFV) promises to revolutionize carrier networks by providing communications service providers (CSPs) with unprecedented flexibility in deploying and managing network equipment and services. But inadequate fault management design can erode the potential cost savings of NFV, resulting in increased maintenance and outage costs.
Wind River recently released a white paper that discusses the benefits and challenges of NFV and discusses how to provide a solid foundation for building carrier-grade NFV network equipment, networks, and services. The paper examines a sample configuration that cuts downtime by up to 97 percent compared to a similar IT-grade solution.
According to a recent study by Ponemon Institute Research, IT data center outages cost an average of $690,240 per incident, or $6,828 per minute. And these IT costs are relatively small compared to public carrier network outage costs: The paper cites a 2011 France Telecom outage that left 28 million customers without phone and messaging service for more than 12 hours, costing the company between $12 million and $25 million in repair costs and customer refunds.
The NFV server is critical to preventing or minimizing the fallout from such failures. Carrier-grade fault tolerance and fault management features can help quickly detect and contain failures, and recover the system to adequately provisioned alternate resources.
The white paper looks at a sample server configuration consisting of eleven Wind River Titanium Server compute servers (ten active and one active-standby) with an IT-grade equivalent solution. Tests predicted just 0.19 minutes of compute server system downtime per year for the Wind River solution, compared to 6.4 minutes per year for the IT-grade solution. A similar gap was found for control system downtime. For carrier deployments, this difference could mean up to tens of millions of dollars in outage costs.
Download the white paper to learn more about the configuration and other ways CSPs can use NFV to significantly reduce CapEx, OpEx, and service introduction times.