Towards Flexibility and Scalability in Parallel Job Scheduling

Gang scheduling has been widely used as a practical solution to the dynamic parallel job scheduling problem. Parallel tasks of a job are scheduled for simultaneous execution on a partition of a parallel computer. Gang Scheduling has many advantages, such as responsiveness, efficient sharing of resources and ease of programming. However, there are two major problems associated with gang scheduling: scalability and the decision of what to do when a task blocks. In this paper we propose a class of scheduling policies, dubbed Concurrent Gang, that is a generalization of gang-scheduling, and allows for the flexible simultaneous scheduling of multiple parallel jobs with different characteristics. Besides that, scalability in Concurrent Gang is achieved through the use of a global clock that coordinates the gang scheduler among different processors.