Virtualization and the cloud have dramatically improved server utilization, but these layers of abstraction have also turned servers and server clusters into amorphous “black boxes” that have made it difficult, if not impossible, to match hardware and software to achieve peak performance. One solution to this problem is to scale data center resources up and/or out as needed, but such brute force scalability is costly, and creates artificially high demands on data center space and power.

A more scalable and cost-effective solution is the Cloud-in-a-Box from Aparna Systems. The Orca μCloud™ has been designed with a system-level approach to server clusters, giving the system an optimal design capable of delivering non-blocking performance and industry-leading price/performance. The patented system architecture has been designed for incremental, in-service upgrades of compute, storage and networking elements, enabling continued increases in cluster-wide price/performance.

The system-level convergence of compute, storage and networking in Aparna’s Cloud-in-a-Box eliminates the storage and network throughput and latency bottlenecks that are at the core of the performance problems in clusters built using blade/rack servers and top-of-rack switches.

The need for compact systems like Aparna’s µCloud was identified in Cisco Global Cloud Index: Forecast and Methodology, 2015–2020: “Network closets and micro data centers are growing in number and importance as internet-connected sensors and devices proliferate and remote users demand faster access to information. In response, organizations will turn to pre-configured micro data center solutions that support fast deployment, greater standardization and remote management across distributed IT locations.”

The ultra-convergence of compute, storage and network resources for peak performance in a configurable, compact server cluster enables Apana’s Cloud-in-a-Box system to deliver a savings of up to 40 percent in capital and operational expenditures, and to reduce the need for rack space and power by up to 80 percent compared to clusters built with blade servers.

Service Provider Applications

  • Network Functions Virtualization (NFV) from the edge to the core
  • Mutli-access Edge Computing (MEC)
  • Fog computing
  • More effective use of space, power and cooling in data centers

Enterprise Applications

  • Private clouds with distributed computing down to the branch office
  • Virtual Network Functions (VNFs)
  • Ultra-converged platform for software-defined storage
  • More effective use of space, power and cooling in central data centers
Robert Noakes