Toward a Common Component Architecture for High-Performance Scientific Computing
HPDC '99 Proceedings of the 8th IEEE International Symposium on High Performance Distributed Computing
A survey of data provenance in e-science
ACM SIGMOD Record
A Framework for Collecting Provenance in Data-Centric Scientific Workflows
ICWS '06 Proceedings of the IEEE International Conference on Web Services
Querying and Creating Visualizations by Analogy
IEEE Transactions on Visualization and Computer Graphics
R2: an application-level kernel for record and replay
OSDI'08 Proceedings of the 8th USENIX conference on Operating systems design and implementation
Provenance collection support in the kepler scientific workflow system
IPAW'06 Proceedings of the 2006 international conference on Provenance and Annotation of Data
Hi-index | 0.00 |
The problem of capturing provenance for computational tasks has recently received significant attention, due to the new set of beneficial uses (for optimization, debugging, etc.) of the recorded data. We develop a provenance collection system aimed at scientific applications that are based on the Common Component Architecture (CCA) that alleviates scientists from the responsibility to manually instrument code in order to collect provenance data. Our system collects provenance data at the granularity of component instances, by automatically recording all method invocations between them, including all input and output parameters. By relying on asynchronous communication and using optimizations to handle large data arrays, the overhead of our system is low-enough to allow continuous provenance collection.