An Adaptive Job Allocation Strategy for Heterogeneous Multiple Clusters

Abstract

In this paper we proposal a new scheduling system based on multi-clusters environments. The work we present is Adaptive Job Allocation Strategy (AJAS) in which the scheduler dispatches jobs that with self-scheduling scheme into appropriate resources across multi-clusters. The strategy focuses on how to increase resource utility, jobs via selfscheduling and dispatching jobs to nodes with similar performance capacities, thus equalizing execution times among all the nodes the jobs require. Experimental results show that the proposed approach performs better than previous schemes.

Symphony: A Scheduler for Client-Server Applications on Coprocessor-Based Heterogeneous Clusters

“…This approach enables the resource sharing at the granularity of frameworks, but does not incorporate application performance requirements such as deadlines and response times in the presence of load spikes. AJAS [27] provides an adaptive job allocation strategy for heterogeneous clusters, but does not consider varying application load and response time to make decisions.…”

Process Rescheduling in High Performance Computing Environments

“…Such an implementation is not extensible, due to the specificity of the shared data structure. Even more, some initiatives use explicit calls in the application code (Bhandarkar et al, 2000) and obligate extra executions to get tuned scheduling data (Silva et al, 2005;Yang & Chou, 2009). A different migration approach happens at middleware level, where changes in the application code and previous knowledge about the system are usually not required.…”