Multicloud computing for on-demand resource provisioning using clustering

Dramatic changes are affecting the world of work which mainly includes the increased impact of information technology. Cloud computing is a technology which helps users to access any shared data or computing services via Internet, with no need to build up an infrastructure to support it. Cloud computing acts as an elastic, flexible and easy way to deploy service platforms using outsourced resources and is therefore gaining a great deal of acceptance in the IT field. The present study focuses on deploying a computing cluster on top of a multi-cloud infrastructure for solving loosely coupled applications. This paper explains the feasibility of the deployment of multicloud infrastructure spanning a local data center (in-house infrastructure) and two different cloud sites, Amazon and Rackspace (external resources). The clustering infrastructure can overcome any failure caused in nodes. Therefore, configuration of the clustered nodes of local data center and the two different cloud sites can improve cost efficiency of the deployment and helps in increasing throughput while solving applications involving many different tasks. As the interfaces to cloud providers are different, a standardized API is highly required, which stands as a limitation. The work assures the credibility of the adopted multicloud system, and it is scalable and flexible in terms of data availability and resource usability.