RASA-A New Grid Task Scheduling Algorithm

In this paper, a new task scheduling algorithm called RASA, considering the distribution and scalability characteristics of grid resources, is proposed. The algorithm is built through a comprehensive study and analysis of two well known task scheduling algorithms, Min-min and Max-min. RASA takes advantages of the both algorithms and avoids their drawbacks. To achieve this, RASA firstly estimates the completion time of the tasks on each of the available grid resources, and then applies the Maxmin and Min-min algorithms, alternatively. In this respect, RASA uses the Min-min strategy to execute small tasks before the large ones, and applies the Maxmin strategy to avoid delays in the execution of the large tasks and to support concurrency in the execution of the large and small tasks. Our experimental results of applying RASA on scheduling independent tasks within grid environments demonstrate the applicability of RASA in achieving schedules with comparatively lower makespan.