Deadline Constrained based Workflow Scheduling for Cloud Computing

Abstract

Cloud computing is the emerging trend in distributing computing that facilitates software applications, platform, and hardware infrastructures as a service. Cloud service providers offer these services based on Service Level Agreements (SLAs) which defines user's required Quality of Service (QoS) parameters. Workflow scheduling is one of the major problems in cloud systems. A good scheduling algorithm must minimize the completion time and cost of workflow application along with QoS requirements of the user. Deadline constrained workflow scheduling model is proposed in this paper to execute workflows with manageable cost and priority while meeting the deadline constraint. Deadline is defined as the latest time by which all the tasks should be completed. The proposed model uses Earliest Finish (EF) to find the deadline and job schedule assignment algorithm to define the paths which is used to schedule the tasks. The proposed model reduces the execution cost and completes the application workflows within the available deadline. This paper extends our earlier work on Artificial Bee Colony algorithm to schedule workflow jobs within the deadline while optimizing the cost and the completion time. The proposed model has been simulated using CloudSim toolkit, and the experimental results show that the proposed algorithm meets the deadline constraint while minimising exhibits less completion time, and also within manageable cost.