Job scheduling strategies for networks of workstations

Abstract

Abstract. In this paper we first introduce the concepts of utilisation ratio and effective speedup and their relations to the system performance. We then describe a two-level scheduling scheme which can be used to achieve good performance for parallel jobs and good response for interactive sequential jobs and also to balance both parallel and sequential workloads. The two-level scheduling can be implemented by introducing on each processor a registration office. We also introduce a loose gang scheduling scheme. This scheme is scalable and has many advantages over existing explicit and implicit coscheduling schemes for scheduling parallel jobs under a time sharing environment.

Notice of Retraction: An improved resource scheduling approach using Job Grouping strategy in grid computing

“…For example Dynamic Scheduling of Process Groups [7] uses hierarchical decision scheduling that is implemented using global and local schedulers. The global scheduler is responsible for long term allocation of system resources, and the local scheduler is responsible for short term decisions concerning processor allocation.…”

Pitfalls in Parallel Job Scheduling Evaluation

“…Some metrics do not describe a real measured value, such as utilization (see pitfall 22). Other metrics may not mean the same thing in different contexts (e.g., slowdown in a time slicing environment vs. nontime-slicing [21,82], or makespan for open vs. closed workloads). Suggestions.…”

Job Re-packing for Enhancing the Performance of Gang Scheduling

“…We have designed a new coscheduling scheme called loose gang scheduling, or scalable gang scheduling [7,8]. Using our scheduling scheme the disadvantages associated with conventional gang scheduling are significantly alleviated, especially the requirement for frequent signal-broadcasting.…”