ABSTRACT
This chapter introduces task scheduling, scheduling challenges, and the classification of scheduling algorithms. It presents several typical heuristic scheduling algorithms, such as Dynamic Level Scheduling (DLS), Modified Critical Path, and Earliest-Finish-Time. The chapter argues that the readers have a preliminary understanding of the workflow scheduling concepts of computing systems resource management. The task scheduling problem is essentially a combinatorial optimization problem. Energy management is essential in modern parallel and distributed systems because various adaptive management techniques are required to maximize energy efficiency and solve thermal problems. Traditional real-time scheduling usually considers scheduling problems from the task level. In centralized scheduling, each task is collected in a master processor unit and then sent to the slave processor unit, where each processor also has an independent scheduling queue. DLS selects the pair with the largest dynamic level for scheduling.
