Placement of software-as-a-service components in cloud computing environment

Abstract

Cloud computing is an emerging paradigm in which information technology resources are provided over the internet as a service to users. Software-as-a-Service (SaaS) is offered by cloud, which can be delivered in a composite form, consisting of a set of application and data components, that works together to deliver higher-level functional software. SaaS components are placed on top of the virtual machines (VMs) in cloud computing environment, which are deployed on physical or storage servers. The SaaS placement is an NP-hard problem. The research problem refers to how a SaaS component is placed on virtual machine to optimize its performance while satisfying the SaaS resource and response time constraints with service level agreement (SLA) constraints. This thesis presents SaaS placement problem as an optimization problem, to maximize the profit of the SaaS providers. Intractability nature of the SaaS placement problem leads to the use of genetic algorithms to obtain sub-optimal solution for SaaS component placement on virtual machine. A suitable codification scheme for SaaS component placement has been proposed for the genetic algorithm. The performance of proposed genetic algorithm has been compared with first-fit randomized algorithm (First-fit RA) by varying number of virtual machines and SaaS components by using in-house simulator. Performance of proposed genetic algorithm has been found to be better in comparison to First-fit RA.