Feature - Cloudbus: A tool for utility-oriented cloud computing

A few of the considerations for users of cloud computing. (Click to enlarge.) Image courtesy Rajkumar Buyya

In order for cloud computing to fully meet its expectations, it needs to overcome its present, somewhat chaotic, challenges. (See figure 3 of full-length paper.) A cloud marketplace — composed of different types of clouds such as computing, storage, and content delivery clouds — will need to be easy  to access and easy to use for both end-users and enterprises.

(We define ‘cloud computing’ here as both the applications delivered as services over the internet, and the hardware and system software in the data centers that provide these services. Both software applications and hardware infrastructures are moved from the private environment to third parties data centers and made accessible through the internet.)

To help do so, our CLOUDS Laboratory at the University of Melbourne, operating under an Australia Research Council grant, has been developing the Cloudbus Toolkit. Our project proposes an architecture for creating market-oriented clouds and computing atmosphere by taking advantage of emerging technologies such as virtual machines (VM).

Our proposed architecture would be comprised of different components working together:  Enterprise Broker, Workflow Engine, Cloud market maker, InterGrid, CloudSim, and Aneka (See attached diagram, currently labeled Figure 6 in full length paper, available here as a pdf).

For example, users can easily access distributed cloud resources for executing their applications using the Enterprise Broker. More complex applications, such as scientific workflows, can be executed using the workflow engine with the support of broker. Instead of directly provisioning cloud resources, users can leave this responsibility to the Market maker, which provides the best options by negotiating with Cloud resource providers.

New architectures offer to unsnarl the problems. (Click to enlarge.) Image courtesy Rajkumar Buyya

Providers, on the other hand, can federate together to build dynamic computing environments that better support the scaling of elastic applications. Large conglomerates can take advantage of our cloud brokering services for building content delivery networks and e-Science applications that can be deployed on infrastructure-as-a-service providers such as Amazon, along with grid mash-ups using VM technologies. New algorithms and setups can be explored in a fully simulated cloud environment provided by the CloudSim framework. For example, energy efficient resource allocation mechanisms and techniques for creation and management of green clouds have been already implemented.

Besides simulation, real deployments of private or public clouds is made possible through Aneka — a platform-as-a-service software system that helps enterprises to build, manage and accelerate their applications. Aneka is being commercialized through Manjrasoft, a spinoff company of the University of Melbourne established in May, 2008. End users can start developing cloud computing applications by using the Aneka software development kit, available for download from the Manjrasoft site.

—Rajkumar Buyya, Suraj Pandey, and Christian Vecchiola. A full-length version is available as a pdf.