Makespan Minimization for Job Co-Scheduling on Chip Multiprocessors
نویسندگان
چکیده
On-chip resource sharing among sibling cores causes resource contention on Chip Multiprocessors (CMP), considerably degrading program performance and system fairness. Job co-scheduling attempts to alleviate the problem by assigning jobs to cores appropriately. Despite many heuristics-based empirical explorations, studies on optimal co-scheduling and its inherent complexity start only recently, and all have concentrated on the minimization of total performance degradations. There is a second important criterion for scheduling, makespan, which determines the finish time of a job set. Its importance for job coscheduling on CMP is increasingly recognized, especially with the rise of CMP-based compute cloud, data centers, and server farms. However, optimal co-scheduling for makespan minimization still remains unexplored. This work uncovers the computational complexity of the problem, and proposes a series of algorithms to either compute or approximate the optimal schedules. We prove that the problem is NP-complete in a general setting. But for a special case (dual-core without job migrations), the problem is solvable in O(n · log n) time (n is the number of jobs). In addition, we present a series of algorithms, including A*-search–based algorithms and a greedy algorithm, to compute or approximate the optimal schedules in the general setting. Experiments on both real and synthetic problems verify the optimality of the optimal co-scheduling algorithms, and demonstrate the reasonable accuracy and scalability of the heuristic algorithms. The findings may advance the current understanding of optimal co-scheduling, facilitate the evaluation of real co-scheduling systems, and provide insights for the development of practical co-scheduling systems.
منابع مشابه
Optimal Co-Scheduling to Minimize Makespan on Chip Multiprocessors
On-chip resource sharing among sibling cores causes resource contention on Chip Multiprocessors (CMP), considerably degrading program performance and system fairness. Job co-scheduling attempts to alleviate the problem by assigning jobs to cores intelligently. Despite many heuristics-based empirical explorations, studies on optimal co-scheduling and its inherent complexity start only recently, ...
متن کاملA note on makespan minimization in proportionate flow shops
We consider two generalizations of the proportionate flow shop scheduling problem with the makespan objective. Each job has a processing requirement and each machine has a characteristic value. In our first (second) case we assume that the time a job occupies a machine is the maximum (minimum) of the processing requirement of the job and the characteristic value of the machine. We show that the...
متن کاملMakespan Minimization for a Parallel Machine Scheduling Problem with Preemption and Job Incompatibility
متن کامل
Some single-machine scheduling with sum-of-processing-time-based and job-position-based processing times
The single-machine scheduling problems with position and sum-of-processing-time based processing times are considered. The actual processing time of a job is defined by function of its scheduled position and total normal processing time of jobs in front of it in the sequence. We provide optimal solutions in polynomial time for some special cases of the makespan minimization and the total comple...
متن کاملA multi Agent System Based on Modified Shifting Bottleneck and Search Techniques for Job Shop Scheduling Problems
This paper presents a multi agent system for the job shop scheduling problems. The proposed system consists of initial scheduling agent, search agents, and schedule management agent. In initial scheduling agent, a modified Shifting Bottleneck is proposed. That is, an effective heuristic approach and can generate a good solution in a low computational effort. In search agents, a hybrid search ap...
متن کامل