PHD Research Proposal for VM Migration in Cloud Computing

The Cloud computing technology [1] uses the pay as go model for sharing the resources such as Virtual Machine (VM), applications, and storage. This computing also provides various benefits such as Self-service provisioning, Scalability, measured service, rapid elasticity, flexibility. The growth rate of Cloud computing technology increases and the demand for Cloud resources are also increasing. Therefore, it is necessary to improve the utilization of resources. However, improper scheduling of the resources leads to resource imbalance, either over-utilization or under-utilization. In order to balance the load of the resources, Virtual Machine (VM) migration [2] is performed, that is the process of moving the VM from one resource to another. In addition, it helps to run the task in the VM even in the failure of its corresponding resources. There are several types of migration. They are cold migration, warm migration and live migration.

Cold migration

  • The cold migration closes the VM on the source resource and begins to execute on the target resource.

Warm migration

  • This migration holds the VM on the source resource for the few seconds, then begin to execute the VM on the target resource without closing the source VM.

Live migration

  • During the execution, the Live migration moves the VM from the source to the target resource, within a second without closing the source VM.

In existing, there are specific algorithms that are available to perform the VM migration through the combination of the utilization threshold strategy and a VM selection policy to propose a power-aware scheduling algorithm THR_MUG [3]. When the VM in the resource gets exceed the threshold level, then the THR_MUG migrates the suitable VM. It minimizes energy consumption and the rate of migration. In the specific research, they assign different algorithm for the dependent and independent tasks for maintaining the load in the resources using a hybrid task scheduling and load balancing scheme in [4]. This scheme uses three different algorithms. Namely, On-Demand scheduling, Querying and Migrating Task (QMT) and Staged Task Migration (STM) that are collectively known as the DeMS. This DeMS [4] successfully minimizes the resource imbalance state by performing the effective scheduling of dependent and independent task. Few of the researchers suggest an efficient VM migration decision-making algorithm with the aim of minimizing the migration in Cloud computing. This technique improves resource utilization; and solves the problems such as load balancing, system maintenance, and SLA violation. Another existing system known as a novel hybrid-copy algorithm in [5] suggests a suitable time for stopping the pre-copy phase and starting the post-copy phase through a parameter in the real-time environment. This system uses the Markov model to predict the memory page accessed order with the aim of analyzing the memory access pattern.

The service providers face difficulties such as SLA violation, energy consumption while providing the service to the users. These difficulties occur due to the resource imbalance, and it is achieved using the VM migration. However, some of the algorithms try to the minimizes the rate of the migration, because VM migration has the capability to increase the energy consumption, and makespan. Also, VM migration is an expensive procedure, since it requires the system resources for managing the abundant VMs, uncertain nature of the workload, and SLA violation. Sometimes, improper VM migration scheduling can affect the performance of nearby applications. Hence, it is necessary to suggest an algorithm with the aim of reducing the migration time, downtime, and SLA violation; and improving the energy-efficiency, resource utilization, VM selection, and target resource selection.

Reference:

  • [1] Alam, Md Imran, Manjusha Pandey, and Siddharth S. Rautaray, “A comprehensive survey on cloud computing,” International Journal of Information Technology and Computer Science (IJITCS), Vol.7, No.2, 68, 2015.

  • [2] Hu, Wenjin, Andrew Hicks, Long Zhang, Eli M. Dow, Vinay Soni, Hao Jiang, Ronny Bull, and Jeanna N. Matthews, “A quantitative study of virtual machine live migration,” In Proceedings of the ACM Cloud and Autonomic Computing Conference, pp.11, 2013.

  • [3] Wu, Xiaodong, Yuzhu Zeng, and Guoxin Lin. “An Energy Efficient VM Migration Algorithm in Data Centers.” 16th IEEE International Symposium on Distributed Computing and Applications to Business, Engineering and Science (DCABES), pp.27-30, 2017.

  • [4] Liu, Yu, Changjie Zhang, Bo Li, and Jianwei Niu, “DeMS: A hybrid scheme of task scheduling and load balancing in computing clusters,” Journal of Network and Computer Applications, Vol.83 pp. 213-220, 2017.

  • [5] Lei, Zhou, Exiong Sun, Shengbo Chen, Jiang Wu, and Wenfeng Shen, “A novel hybrid-copy algorithm for live migration of virtual machine,” Future Internet, Vol.9, No.3, pp.37, 2017.

Leave Comment

Your email address will not be published. Required fields are marked *

clear formSubmit