Abstract—In a cloud computing, provisioned virtual machines are insufficient for computing the services it needs to be fine tuned by adding the additional resources or boosting the existing resources. A static method of resource allocation becomes inefficient under the different load. Based on the heavy load virtual machine can be adopted for dynamically changes of its behavior. There are two ways of scaling in the cloud scenario as horizontal scaling and vertical scaling. In a horizontal scaling, if the virtual machine resources are suffer from providing a service a new virtual machine was created and launched immediately. A threshold value has been maintained for scale out and scale down the virtual machine. Horizontal scaling is a traditional and well suitable approach for cloud computing environment. But the limitation of this approach is required separate load balancer to distribute the load between the virtual machines. In a vertical scaling resources are boosted by maximizing the virtual machine capabilities without shutting down the virtual machines. It’s not like a traditional scaling approach to set the threshold and adding the resources. Here the resources are monitored by certain interval and based on the analysis resources are added to existing virtual machine. But lots of challenges are there in this approach. The limitation of this approach is we can’t scale up the virtual machine up to the physical machine capability. Here we presented and analysis our approach for increasing the CPU capability of virtual machine. Likewise we can increase other resources like memory, vDisk and bandwidth. presented architecture resides on top of the virtual machine monitor and acts based on the scheduling algorithm.
Index Terms—Cloud computing, vertical scaling, horizontal scaling.
The authors are with Center for Development of Advanced Computing, Chennai, India (e-mail: {sakthis, mahendrane, mehanathenn}@cdac.in).
[PDF]
Cite: Sakthi Saravanankumar P., Mahendran Ellappan, and Mehanathen N., "CPU Resizing Vertical Scaling on Cloud," International Journal of Future Computer and Communication vol. 4, no. 1, pp. 55-58, 2015.
