Date of Publication :8th November 2017

Abstract: Resource pooling in cloud is considered as a shared pool of computing resources such as Cores, RAM, Storage, Disk etc. It helps in the resource allocation among users, thus increasing the utilization rate and saves cost. There are many large shared cloud resource pools such as Amazon EC2 cloud, Google Cloud etc which monitor the usage of computing resources and generates the billing to users based on the usage of resources. These are commercial public cloud and resource pooling implementation is done using cost-optimization approach. In this paper, we deployed a resource pool of opensource private cloud openstack using MAAS (Metal as a Service) and Juju. There are many opensource cloud platforms such as Openstack, Eucalyptus, Opennebula etc for deployment of private cloud. For production use, a openstack tool, MAAS (Metal as a Service) with Juju manages the computing resources more flexibly. First, the MAAS ubuntu server cluster is deployed on a private network in which machine Cores, RAM, Storage, Disk, etc are added dynamically using pre-defined powertype. Furthermore, Juju environment is configured and Juju Bootstrapping is done for assignment of openstack services. The pooling configuration is done step-by-step and a resource pool of six machines is successfully deployed, each machine lists the number of Cores, RAM, Storage, Disk, etc. As a result, the openstack services are assigned to deployed machines, thus allocation of resources would be easily monitored and controlled.

Reference :