The Cloud Scheduler Politics of Multiprocessor Multitask Real Time Systems
Song Mei Yuan · Chinese Journal of Computers · 2000
The scheduler of multiprocessor among tasks is a key technology to massively parallel processing (MPP) and real time systems.This paper analyses the importance and operating timing of tasks with cloud method,evaluates the value and elapsed time of activated tasks,and puts forward a new scheduler policy,cloud scheduler policy, which can take both the value and elapsed time of tasks into account.Cloud scheduler policy may solve the problem of how to define task PRI and task operating timing well,improve the using efficiency of system resources,operating quality,and enhance the system effectiveness.By means of simulation,it is proved that cloud scheduler policy is better than PRI and FIFO scheduler policies that are very widely used today.