Abstract

Federated cloud has emerged as a cost-effective and scalable platform for multiple cloud providers to collaborate in serving the growing demand for cloud resources. In a federation, the major concern for each cloud provider is to maximize its own profit while meeting the quality of service requirements of end-users. Virtual machine (VM) migration allows the cloud provider to rent resources across the other cloud providers in a federation to minimize the cost. However, VM migration comes with an additional penalty in terms of time and cost, and it might degrade the performance of the system during migration.We address the problem of cost-aware VM migration in federated clouds to maximize the profit of the cloud provider by minimizing the cost. We analyze the impact of VM migration across the data centers in a federation by modeling migration cost, migration time, and downtime. We formulate an optimization problem to minimize the total cost of operation (TCO) of the cloud provider as well as the migration time. We model the TCO to include the cost of hosted VMs and their migration cost. We propose efficient heuristics to select the source and destination data centers for migration, as well as the VM to be migrated. Based on these heuristics, we design a polynomial-time VM migration algorithm to minimize the TCO and migration time. The proposed algorithm considers the characteristics of the VMs, the bandwidth and the data transfer price across data centers to give the highest benefit in migration. We evaluate the proposed algorithm using CloudSim across different scenarios. Our results demonstrate the advantages of VM migration in a federated cloud for both the cloud provider and the end-user, to lower the TCO and the migration downtime. The proposed algorithm is shown to outperform other baseline approaches, with up to 48% reduction in the TCO along with reducing the migration time, downtime, and number of migrations. We also evaluate the impact of network bandwidth and the page dirty rate of VMs on migration performance.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.