L. Schley (Germany)
Grid Computing, Resource Management, Co-Allocation, Data Management, Scheduling
Currently, middleware systems for Grid computing like gLite do not integrate knowledge about data availability into the scheduling process. That is, data may be on tertiary storage, e.g. hierarchical storage management, and first ac cess is delayed. To this end, we investigate the gLite mid dleware software stack and highlight those services which could be enhanced to increase system utilization. Here, we focus on modifications of the workload management sys tem and the computing element. As the computing element depends on a resource management system, we discuss the impact of file systems on the scheduling process. Optimiz ing the LHCb job workflow, we design a scheduling strat egy which respects those constraints. Conceptually, this strategy performs a co-allocation of data movements con currently to computational processes. Based on a workload trace from the Parallel Workload Archive we are able to present improvements in terms of system utilization and average weighted response time.
Important Links:
Go Back
