FOR 1306/1: Stratosphere II – Information Management on the Cloud (TP D)

Processing iterative, stateful data analysis programs on vast amounts of “data in motion” under low-latency requires novel methods and techniques both from systems and an algorithmic point-of-view. In the proposed second phase of the research unit, we plan to build on our existing results to considerably advance the state-of-the-art in designing and building systems for executing complex data analysis programs on potentially evolving datasets under the constraint of significantly reducing latency (i.e., the time until first analysis results are available). We will use the existing Stratosphere I system with its runtime system, compiler, optimizer, and scripting languages as a foundation for implementing our ideas.