Residue Based Adaptive Resource Provisioning through Multi-Criteria Decision and Horizontal Scaling
Pradeep Kumar Vadla · International Journal of Advanced Trends in Computer Science and Engineering · 2020
Elasticity and scalability are prominent issues in cloud computing which are resolved effectively using federated clouds.The agent-based model is simulated in our work in which all the elements of cloud computing are categorized into specific agents like cloud consumer agent, cloud provider agent and cloud broker agent.The collaborated cloud providers who are contributing resources are treated as collated cloud provider agents.The residue-based resource provisioning is carried at cloud broker by performing multi-criteria decision for finding dominant collated provider agent in providing resources within the limit of service level agreement and horizontal scaling of the virtual machine is done based on greedy cloud ranker algorithm to rank the cloud which contributes the virtual VM which satisfy the consumer agent request within specific turnaround time without violating service level agreement.The features of the interoperable cloud are simulated using python classes and realization of horizontal scaling is tested for computing percentage of request satisfaction with full or partial and transaction rate completion of allocating virtual machine to cloud consumer agent.