To ensure the correctness of the grid workflow specification and execution in grid computing environments, this project aims to develop effective and efficient verification methods and algorithms, and dynamic adjusting mechanisms for the inconsistency removal without stopping and replacing the grid workflow execution.
Background
Grid workflow systems which are evoking a high degree of interest aim to support modelling, redesign and execution of large-scale sophisticated multi-institutional e-science and e-business processes using the grid computing approach to enable the coordinated use of numerous distributed and heterogeneous resources . Currently, not much research is being done and mainly focuses on the functional aspects like grid workflow specification languages targeting grid applications, grid workflow execution engines which interoperate with grid services, grid services that support grid workflow execution, graphical grid workflow interface systems. However, the corresponding non-functional aspects such as grid workflow management, grid workflow QoS control, grid workflow security are still open.
One of the important open aspects of the grid workflow management is grid workflow verification and subsequent inconsistency removal. It is clear that during the grid workflow specification and execution, there could be some inconsistencies existing such as some exceptions and errors. Therefore, on one hand, it becomes very important how to verify the grid workflow specification and execution to identify the possible inconsistencies. On the other hand, only identification of the possible inconsistencies is not the ultimate goal, and we must develop some mechanisms to remove the inconsistencies, especially in dynamic OGSA-compliant grid computing environments without stopping and replacing the grid workflow execution. In our view, the primary grid workflow verification research should focus on grid workflow structure verification, grid workflow resource verification and grid workflow temporal verification. Grid workflow structure verification aims to identify the structure inconsistencies such as deadlock, livelock, dead activities and so on. Grid workflow resource verification aims to identify the resource inconsistencies such as resource performance, budget & cost conflicts, resource allocation conflicts and so on. Grid workflow temporal verification aims to identify the temporal violations since in the real world, any complex e-science or e-business process grid workflow systems are time constrained.
Research Scope
The research of the project mainly includes the following key issues:
Grid workflow representation for verification
Effective and efficient methods and algorithms for grid workflow structure verification
Dynamic removal mechanisms for removing the corresponding inconsistencies without stopping or replacing the grid workflow execution
Simulation and evaluation of the developed methods, algorithms and mechanisms
