The ASAP FP7 research project developed a dynamic open-source execution framework for scalable data analytics. The underlying idea was that no single execution model is suitable for all types of tasks, and no single data model (and store) is suitable for all types of data. Complex analytical tasks over multi-engine environments therefore require integrated profiling, modeling, planning and scheduling functions.
- A generic task-parallel programming model in conjunction with two runtime systems for distributed or parallel execution in the cloud. The runtimes include state-of-the-art features such as irregular general-purpose computations, resource elasticity, synchronization, data-transfer, locality and scheduling abstraction, the ability to handle large sets of irregularly distributed data, and fault-tolerance.
- A modeling framework that constantly evaluates the cost, quality and performance of available computational resources in order to decide on the most advantageous store, indexing and execution pattern.
- An adaptation methodology to enable analytics experts to amend submitted workflows by changing or modifying a workflow while it is being processed. Users can change the parameters of operators already comprised in the workflow, or the structure of the workflow by removing or adding operators.
- A visual analytics dashboard to show query results and metadata in an intuitive manner, with special focus on the interactive exploration of datasets, dynamic temporal controls, on-the-fly query refinement mechanisms, and the geospatial projection of structured and unstructured data.
The generic nature of the ASAP architecture supports a wide range of different tasks. Within the project, the consortium focused on the real-time analysis of Web content and telecommunications data.