An Efficient Fully Dynamic Algorithm to Optimize the Task Scheduling Activities in Cloud Systems Sarwat Jahan 1 , Safdar Tanweer 2 , Masoorul Haque 3 1 Research Scholar, M. Tech. (CS), Jamia Hamdard, New Delhi, India 2 Assistant Professor, Dept. of Computer Science, Jamia Hamdard, New Delhi, India 3 Principal System Analyst, National Informatics Centre, Govt. of India, New Delhi, India Abstract— The model of cloud computing has been evolved as a very popular and interesting model that concentrates on the on demand services very efficiently whenever they are needed by cloud terminals. In a quite different computing style, the delivery of massively scalable resources to the end users/clients/customers is carried out ‘as a service’ within the cloud systems using internet as a communication channel. In the cloud systems, the selection of the best suitable resources for the execution of arriving tasks is decided by the used scheduling strategies after considering some static and dynamic behaviour and the restrictions applied on them. From the users’ point of view, some other issues like task execution cost or task completion time may become essential parameters for deciding the scheduling algorithm efficient. The service providers always try to provide the resources in optimized manner so that the utilization of resources may be optimum and the resource potential might be left minimum. This paper proposes an efficient scheduling algorithm by which these big challenges may be addressed efficiently in the cloud systems. The arriving tasks would be bound after measuring them on the basis of their requirement like minimum task execution cost or minimum task completion time and their priority levels. The selection of the resources would be made on the basis of greedy approach based task constraints. This proposed work has been implemented, tested, validated and verified on a simulator. Results Show the effectiveness and correctness of the proposed framework and come out with a significant improvement over sequential scheduling algorithms. Keywords— Cloud Computing, Task Scheduling, Optimal Scheduling, Sequential Scheduling, Dynamic task execution. I. INTRODUCTION Cloud computing is one of the most current emerging computing technologies and has received a lot of interest by the researchers now a days. It can be explained as on demand pay-per-use model which provide the software, information and resources in a shared mode as per the users’/clients’ requirements when the need for it [1]. It is evident that the human dependency has been focused on the cloud computing for the last decade. The most popular online gaming sites, document sharing, social networking, email hosting and business sites are moving to cloud environment from their traditional computing environment. Google, IBM, Microsoft, Yahoo, Apple and Amazon are the famous initiators in this field. The theory of scheduling is taking a great attention of researchers with increase in its popularity. Basically the term scheduling can be defined in the context of cloud computing as ‘It is the appropriate mapping process in which the arriving tasks are mapped to the available cloud resources on the basis of the requirements and characteristics associated with the tasks.’ The scheduling is the very important aspect for the appropriate, efficient and effective working within the cloud systems. The various task attributes/parameters need to be considered while designing efficient and optimized scheduling. The various cloud resources must be fully optimized by used scheduling algorithm without tempering the pre-defined service guidelines/parameters of existing cloud system. Generally, the process of scheduling is carried out within the cloud systems into three stages as shown in Figure 1:- Fig 1: Scheduling Process  Resource discovering and filtering– The discovering and filtering of the resources is the responsibility of Datacenter Broker. Datacenter Broker also collects and updates the information related to the filtering and resource discovery.  Resource selection– The selection of the target resources is finalized on the basis of certain task parameters of the arriving tasks.  Task submission- In this stage, the tasks are submitted to selected resources. The above mentioned scheduling steps are further Sarwat Jahan et al, / (IJCSIT) International Journal of Computer Science and Information Technologies, Vol. 6 (2) , 2015, 1328-1332 www.ijcsit.com 1328