DSMS Study - Study on Real-time Stream Processing Engines

Institutions

• Computer and Communication Systems, University of Innsbruck
• Data Management, University of Erlangen

Team

• Univ.-Prof. Dr.-Ing. Falko Dressler (PI)
• Prof. Dr.-Ing. Klaus Meyer-Wegener (PI in Erlangen)
• Dr.-Ing. Michael Daum
• Niko Pollner
• Frank Lauterwald

Funding

• Siemens AG

Project Time

• 01.07.2011 - 31.09.2011

Description

The goal of this study is the investigation (evaluation) of Data Stream Management System (DSMS) engines for the purpose of real-world applications. The main diffences of existing engines are to be identified and described. The feasibility for particular application scenarios shall be named. The emphasis is on restrictions and the potential for adaptation with respect to the following aspects:

• input streams of varying complexity,
• query language,
• user interface and development tools (input of new queries, test and debugging tools), and
• access to stored data.

In a second part, the study should describe for a smaller number of engines the actual implementation of a given scenario. This scenario is a simple version of airport vehicle tracking, with warnings to be issued when one of the vehicles enters a forbidden area. A demonstration is to be built with two diffent engines that uses some artificial input data and shows the execution of several queries.