Resource Management in Cloud Computing Environment

Abstract

Cloud Computing is a very dynamic environment and it uses a pay per use model. An effective management of cloud resources is advantageous to Cloud Service Providers and Cloud Service Users both. One of the main problems to be answered at the service provider side is, to achieve the efficient utilization of cloud resources without violating customer Service Level Agreements (SLAs). Similarly, cloud users need very dynamic resource provisioning mechanism for their cloud based applications to reduce service charges (Auto Scaling). In this research, We have suggested three policies for SLA aware Auto scaling and one policy for server consolidations. A private cloud is set up using the Open Nebula Framework and KVM as hypervisor. In this setup, we executed several experiments like VM live migration, application load balancing and scaling etc. Along with that, we have used the Cloudsim simulator to simulate our auto scaling, consolidation and load balancing policies. Experimental results in simulation environment show that effective auto scaling policies are useful for dynamic resource provisioning. The suggested consolidation approach can be used by cloud service providers to perform server consolidation with lesser number of SLA violations. newline newline