Efficient algorithm for workflow scheduling in cloud computing environment

Abstract

Executing a large number of workflow applications within their deadlines and efficient utilization of computing resources in a cloud computing environment is a challenging problem. A workflow application is usually represented as a set of tasks interconnected via data. In most of the scheduling algorithms, the execution times of the tasks are pre-computed. However, the execution time of the tasks is computed based on the availability of computing resources. On the other hand, offering flexible and elastic computing resources can handle a large number of applications in order to utilize the resources efficiently and maximize the revenue generation. In this paper, we propose an efficient workflow scheduling algorithm (EWSA) which can handle a large number of applications simultaneously. The objective of the algorithm is to estimate the execution time of all the tasks dynamically. The algorithm also creates a suitable VMs with minimum resources such that the entire application can be executed within its deadline. Through simulation, we establish that the proposed algorithm performs better than the existing algorithm in terms of various performance metrics.