Seamless event and data stream processing: Reconciling windows and consumption modes

Abstract

For a number of stream applications, synergistic integration of stream as well as event processing is becoming a necessity. However, the relationship between windows and consumption modes has not been studied in the literature. A clear understanding of this relationship is important for integrating the two synergistically as well as detecting meaningful complex events using events generated by a stream processing system. In this paper, we analyze the notion of windows introduced for stream processing and the notion of consumption modes introduced for event processing. Based on the analysis, this paper proposes several approaches for combining the two and investigates their ramifications. We present conclusions based on our analysis and an integrated architecture that currently supports one of the reconciled approaches. © 2011 Springer-Verlag.