Monitoring data dependencies in concurrent process execution through delta-enabled grid services

This paper presents our results with monitoring data dependencies among concurrently executing, distributed processes that execute over Grid Services. The research has been conducted in the context of the DeltaGrid project, focusing on the development of a semantically-robust execution environment for the composition of Grid Services. Delta-Enabled Grid Services (DEGS) are a foundational aspect of the DeltaGrid environment, extending Grid Services with the capability of recording incremental data changes, known as deltas. Deltas generated by DEGS are forwarded to a Process History Capture System that organizes deltas from distributed sources into a global, time-sequenced schedule of data changes. The design and construction of Delta-Enabled Grid Services is presented, along with storage and indexing techniques for merging deltas from multiple DEGS to create a global schedule of data changes that can be analyzed to determine how the failure and recovery of one process can potentially affect other data-dependent processes. The paper also summarizes the performance results for the global history construction and retrieval process.