International Conference and workshop on Advanced Computing 2013 |
Foundation of Computer Science USA |
ICWAC - Number 1 |
June 2013 |
Authors: Ruta Ajgaonkar |
f5493f32-0775-4a7d-a3c6-7e3a63b33a9d |
Ruta Ajgaonkar . Hybrid Approach for Optimal Provisioning in Cloud. International Conference and workshop on Advanced Computing 2013. ICWAC, 1 (June 2013), 0-0.
Cloud Computing is a paradigm where services are made available through the internet. It is providing services using many technologies. It has gained importance because of its pay-per-use structure which means the user has to pay only for his use. Because of the large number of cloud users, cloud Management has become of prime importance. Cloud Management involves many tasks such as provisioning, deployment, monitoring, metering, billing etc. To provide cloud computing services, provisioning of the cloud resources is a prior step. Cloud resources include CPU, memory, disk, network bandwidth etc. Provisioning is simply allocation of resources. To provide services to the user or to deploy applications on the cloud, provisioning of resources has to be done. Due to the availability of limited resources in the cloud, a need for optimal use of resources is felt. Optimal allocation of resources, involves conservation of resources, minimizing the cost of resources, satisfying the need of the user etc. Minimization of cost for renting virtual machines is considered, because cloud has gained popularity due to its cost effective nature. The cost of renting a virtual machine is calculated by considering the cost of RAM utilized, CPU cores rented, the bandwidth utilized, also considering the time for which the virtual machine is rented. The focus of this project is on minimizing the cost of RAM utilized which in turn minimizes the cost of renting the virtual machine. Optimal allocation of resources can be done by cost minimization, load balancing, cost balancing etc. A Hybrid approach for optimal allocation has been developed to minimize the cost of the virtual machine that is rented by the user.