As network functions virtualization (NFV) has gained credence over the past few years, network operators are embracing the idea of separating application logic from the underlying proprietary hardware platforms used to run telecom applications in favor of commercial off-the-shelf (COTS) hardware deployed as a cloud infrastructure, on which these functions can run as virtual machines or containers.

The flexibility and cost savings from this approach are significant. The next generation services that will be delivered require data (and infrastructure) to reside at the edge of the network, as opposed to being delivered from a centralized cloud. Consequently, the ability to deliver cloud infrastructure on top of COTS hardware in smaller edge environments is key to achieving these benefits.

The industry has directed much of its attention on OpenStack and its offshoots, Akraino and StarlingX, to try and address these edge use cases. The challenge is that OpenStack was designed for building massive hyperscale clouds, and the problem is that they are now trying to fit it into a use case for which it was not designed.

OpenStack’s architecture, management/deployment complexity, and large footprint are all non-issues when talking about a large, centralized, cloud infrastructure. However, they become problematic when talking about multitudes of highly distributed nodes, some of which need to reside inside a small street cabinet or on a desk at a customer’s premises – places where a minimum 6 node cluster of rackmount servers with 64GB of RAM are simply not feasible.

NodeWeaver was designed to deliver cloud functionality at the edge:

• Software only and runs on virtually any x86-based hardware
• Lightweight codebase enables it to run even on SoCs like Intel Atom and AMD Epyc and Ryzens
• Full reliability on as few as 2 nodes
• Autonomic management features reduce manual intervention
• Create clusters with heterogeneous hardware

Customer Spotlight:

One of the world’s leading cruise lines is using NodeWeaver as the foundational platform for a global distributed edge infrastructure solution at all ship and shore-based operations to support a VNF rollout across their enterprise.

By virtualizing multiple network functions and consolidating them onto a single platform, system resiliency is critical to maintaining network services. The ability to create flexible, scalable, and highly available nano-clouds at the edge, which autonomically addresses failures as well as the ability to easily manage hundreds of distributed locations around the world with minimal human intervention were the key criteria for this project.

After evaluating other uCPE platforms and solutions, they selected NodeWeaver due to its simplicity, flexibility, reliability, and time to value.