Efficient categorization placement and migration of virtual machines in cloud sdn environment

Abstract

Cloud Computing has emerged as a successful technological newlineparadigm by providing successful and on demand provisioning of resources newlinesatisfying a large number of customers. This provisioning is done with the newlinedeployment of very large data centers with multiple servers taking advantage newlineof the virtualization technology. These data centers deployed with multiple newlineserver leads to high energy consumption, communication and operating costs. newlineThe design of a virtualized data center requires efficient allocation of resources newlineto meet the customer requirements. As the services are offered through Virtual newlineMachines (VMs), efficient management of the Virtual Machines is the key to newlinethe success of the Cloud Computing technological paradigm. newlineThe quality and the resiliency of the services offered by any Cloud newlineService Provider (CSP) has gained its popularity and thus play a vital role in newlinethe current situation. With the advent of Software Defined Network (SDN) newlinetechnologies, great improvements are achieved in handling services offered by newlinethe cloud service providers as well as in the underlying cloud infrastructure. At newlinethe time of offering the services by the cloud computing environment, the newlineperformance parameters such as the service delay and jitter can be reduced by newlineallocating sufficient bandwidth. This is possible with the knowledge of the newlineunderlying cloud infrastructure. SDN technology has a centralized control over newlinethe underlying cloud infrastructure and thus greatly improves the efficiency of newlineservices through proper management of VMs. newlineIn this thesis, issues related to the Allocation, Placement and newlineMigration of VMs are addressed. The methods, techniques and algorithms used newlinefor Virtual Machine Placement (VMP) aims to improve the resource utilization newlineand reduce the communication cost. This thesis provides both static and newlinedynamic VMP algorithms to improve the process of VMP. newline