Scalable Parallelization of Expensive Continuous Queries over Massive Data Streams
Author: Dr. Erik Zeitler
Language: English
Affiliation: Uppsala University, Sweden
Abstract:For applications that require execution of non-trivial Continuous Queries (CQs) over data streams of high rate, the execution of the CQs must be parallelized. Declarative stream splitting functions are introduced in this Thesis, allowing the user to specify customized parallelization in the query language, similar to fragmentation and replication of distributed databases. It is shown how to automatically generate parallel execution plans for stream splitting functions that enable CQs to be processed at stream rates close to network speed.
Type of work: PhD
Area: Big Data: Analytics, Data Stream Management Systems
No of Pages: 152
Year of completion: 2011
Name of supervisor/affiliation: Tore Risch/Uppsala University, Sweden