On the Necessity for High-availability Data Center Backends in a Distributed Wireless System

Abstract

When business processes depend on the processing capabilities within a data center, the typical system architecture use a high-availability setup to maintain a high level of service. Faced with a specific machine-to-machine system consisting of many endpoints that collect and forward data to the data center we argue that the dependability of the overall system does not necessitate a high level of service for the data center components. Taking an existing discrete event simulation model of a distributed technical system we investigate and discuss the effects of prolonged outages of the data center on the major business processes of the system.