Resource and service orchestration for deploying openstack cloud using MAAS and Juju

Abstract

There is a huge demand for cloud computing platforms as they are a key enabler of digitalization. OpenStack is one of the prevalent cloud computing platforms in the industry. It is an open-source project which is gaining a lot of traction known for handing the core cloud-computing services of networking, storage, identity, and image services. There are several other open-source cloud technologies which are also available for private cloud implementations. However, we chose OpenStack due to its popularity, flexibility, and stability. The current manually based installation of OpenStack is very complex, tedious, and time-consuming, and therefore not ideal for a fast-paced environment such as high demand data intensity ecosystems, which are at the core of the current digital transformation. The primary objective for a cloud administrator is to commission and deploy OpenStack as fast as possible. Thus, in this paper, we used MAAS (Metal as a Service) and Juju to orchestrate the deployment of OpenStack cloud services. In the process of deploying OpenStack, the MAAS Ubuntu server cluster is first deployed on a private network, where machine Cores, RAM, Storage, and Disk are dynamically added using pre-defined power types. In addition, the Juju environment is set up, and Juju bootstrapping for the assignment of OpenStack services is completed. Step by step, five machines enlist on MAAS and display their resources until they are successfully deployed, with each machine listing the number of Cores, RAM, Storage, Disk, and so on. As a result, the OpenStack services are assigned to deployed machines, allowing for easy monitoring and control of resource and service orchestration.