Lookup system design for networked cloud

Ever increasing numbers and attributes of the hardware, software, and contents require robust lookup system that scales with the size and complexities of services and applications of Future Internet. The lookup function, as a fundamental building block of network system, becomes more crucial than ever to ensure high quality of services. In this paper, we investigate a lookup system design using replicas and caches in a networked cloud environment. The proposed design would then be evaluated in terms of key requirements such as query response time, bandwidth, and storage. A logical analysis and visualization of data sets are discussed to support our design decisions. The lookup system which meets performance requirements, given by a service provider, is modeled as an integer linear programming problem. We discuss our solution approaches, including a newly proposed heuristic method demonstrating reasonable performance and near optimal solutions, and impact and tradeoffs of various design parameters.