Our proposition for prevailing over the various challenges and achieving the full potential of Big Data and enabling the IoT evolution is a universal architecture and software ecosystem. The embodiment of this novel universal ecosystem is a new class of server software we refer to as UniServer.
The use of UniServer enables the following salient attributes:
- high performance able to satisfy quality of service requirements of any Internet Service,
- easy to deploy to both traditional cloud data-centres as well as to data-centres in a box deployed at the edge of the cloud,
- highly energy efficient to reduce the capital and operating costs of traditional cloud data-centres but also to facilitate the utilization near the sources of data using a home power distribution network without the need for expensive and complicated cooling infrastructure,
- supporting an easily programmable system software stack consisting of a virtualization environment that can seamlessly combine fog resources with the resources in a backbone cloud,
- process the data reliably and securely without compromising their privacy and potential confidential nature.
To realize the above UniServer attributes we have set out the following main objectives:
- Objective-1 Enable energy efficient operation at the edge of the cloud
- Objective-2 Shave the safety margins by exposing intrinsic heterogeneity
- Objective-3 Develop error resilient system software for exploiting operation at extended margins
- Objective-4 Demonstrate the potential and explore UniServer trade-offs for various cloud and edge applications