Deployment Approaches in Distributed Complex Event Processing

Abstract

Big Data technologies have traditionally focused on processing human-generated data, while neglecting the vast amounts of data generated by Machine-to-Machine (M2M) interactions and Internet-of-Things (IoT) platforms. These interactions generate real-time data streams that are highly structured, often in the form of a series of event occurrences. In this paper, we aim to provide a comprehensive overview of the main research issues in Complex Event Processing (CEP) techniques, with a special focus on optimizing the distribution of event handlers between working nodes. We introduce and compare different deployment strategies for CEP event handlers. These strategies define how the event handlers are distributed over different working nodes. In this paper we consider the distributed approach, because it ensures, that the event handlers are scalable, fault-tolerant, and can handle large volumes of data.