Abstract
Analysts have marked Event Processing as the most growing segment in enterprise computing during years 2008 and 2009, furthermore, this trend is expected to continue. Many of the large and medium software companies (IBM, Oracle, Microsoft, Sybase, Progress Software, Software AG and TIBCO) are now offering event processing products as well as a collection of smaller companies. Other indications for the emerging nature of this area are: extensive coverage by analysts as a separate area, the establishment of a dedicated research community with an annual conference (DEBS), and the establishment of a consortium that includes vendors and academic people as EPTS (Event Processing Technical Society) http://www.ep-ts.com/. The early event processing commercial products were mostly descendents of research projects rooted in multiple disciplines, some of them are data management disciplines (active databases, data stream management, temporal databases) and some are rooted in other areas (discrete event simulation, distributed computing, formal verification). The tutorial is intended for a technical audience that is interested in deep dive into understanding event processing. The audience will gain insights about event processing: What it really means? Where does it come from? How does it relate to research concepts (e.g. stream computing) as well as enterprise computing terms (e.g. Business Rules Management Systems)? The audience will also gain insights into the current state of the art, the leading architectures, the basic building blocks of event processing, and the various programming styles exemplified by code examples. Last but not least, the audience will gain insights about the current trends, and the research challenges that exist, this part will be based on the discussions in the Event Processing Dagstuhl seminar that was held in May 2010 http://www.dagstuhl.de/de/programm/kalender/semhp/?semnr=10 201. The tutorial slides are available for public viewing on slideshare http://www.slideshare.net/search/slideshow?q=opher+etzion+%2B+vldb2010+tutorial The current generation of event processing products [14] has been preceded by several research projects in the 1990-ies: Rapide in Stanford [15], Infospheres in Cal Tech [2], Apama in Cambridge University [6] and Amit in IBM Haifa Research Lab [1]. In later phase there were some research project that have taken the stream oriented approach such as the Stanford Stream project [3] and Aurora [5]. A collection of start-up companies, many of them descendents of these projects have emerged, some survived, and some did not. From those who survived some were acquired by bigger companies. Event processing as a research discipline has multiple ancestors. In the database area it is a descendent of work done in active databases [19], temporal databases [12], and data stream management systems [9]. Other ancestors are inference rules, discrete event simulation, and distributed computing (pub/sub). © 2010 VLDB Endowment.