The emergence and use of new technological solutions such as cloud computing is due to the rapid evolution of business processes, the need for data storage, the sharing of services and the miniaturization/substitution of machines and tools for work/calculation [1]. The cloud makes it possible to store information and focuses on the data regardless of their medium. The internet allows the cloud to deliver services anytime and anywhere (location), which allows for ubiquitous services. Cloud computing is a new way for businesses to buy and consume services related to information systems (IS) around the world via the Internet. The openness and ease of use of the Internet makes computer networks accessible to all other users of the network (Internet). It enables virtualized applications, software, platforms, compute, and storage to be quickly provisioned, scaled, and released instantly [2].

Cloud is used for solving complex data intensive applications [2], whereas workflows are used to coordinate data intensive applications in many different scientific disciplines. Cloud computing allows wide computation options along with resource facilities for executing of various workflow applications. Scientific workflows are largely accepted as a useful paradigm to describe, organize, and distribute complex scientific analysis.

Workflow technologies are integrated through resource provisioning technologies in Cloud so as to determine the proper quantity of resources required for any execution of workflows [3, 4]. This is done to reduce the financial cost from user perspectives and to fully exploit the resource utilization from cloud providers. Workflow Managements Systems (WMS) are charged for defining, creating and managing the execution of these workflows. In order to effectively schedule workflow tasks on cloud environments, a WMS uses an efficient scheduling method for allocating the tasks in a workflow to appropriate cloud resources with the aim of satisfying user requirements, well known as Quality of Services (QoS) constraints.

As this problem is a NP-complete and depends on the problem size [3], abundant scheduling algorithms have been proposed to resolve challenging problems quicker than metaheuristic one [4].

In addition to the usual QoS constraints such as time and cost to solve this problem, energy consumption has turn into a key concern in the field of Cloud computing.

This paper aims to implement workflow scheduling using Dynamic Provisioning Based on Demand (DPBD) algorithm. Minimizing makespan and energy consumption of the cloud service are the dual objectives, whereas completing the tasks in a sequential manner and the priority of the tasks are the design constraints. The Pareto optimal solutions of the multi-objective optimization problem are obtained using the non-dominated sorting genetic algorithm (NSGA-II).