The Design and Implementation of a Semi-autonomic Data Stream Query Processing Testbed

This work describes the design and implementation of data stream query processing system built with two goals in mind. The main goal is to create a single-user data stream query processing system designed for small to medium scale live-data applications and for querying large datasets. The system will have features such as minimal setup and configuration, an easy-to-use yet powerful query language, autonomic buffer management, semi-autonomic data discard policies, and the ability to efficiently process queries for large, structured data files. The second goal is to create a testbed for researching current problems in data stream query processing. Some areas of interest include buffer management techniques, data discard policies, and large dataset query processing. In order to build a system that is both useful and efficient, we have built our query processing engine using a modular design that utilizes various design patterns. This allows us to design and test components independently and analyze them at a granular level.