Abstract & Overview
Cloud computing is such a revolution concerning the IT world offering computing as services capable of diminishing operational costs and complications. Recently, these service models, ranging from IaaS, PaaS, and SaaS, and deployment models in private, public, and hybrid clouds, offer users almost unlimited computing and storage capabilities on a pay-per-use basis. This elasticity of cloud systems makes it very easy to dynamically provision and de-provision resources to cater to very different needs. This facility has led to its widespread use within domains such as social networking, defense, scientific computing, financial services, and medical. IDG Communications has now announced that 73% of corporations are currently utilizing clouds, with a further 17% in the process of implementing. Service abstraction to increase usability raises yet a fresh set of issues in terms of operational costs, reliability, energy efficiency, and security. Especially in cases where the framework is applicable to critical ventures, as exhibited just a while back by Knight Capital in 2013, system failures may have serious financial and credibility repercussions. Fault tolerance strategies through resource redundancy increase the cost of downtime risk but lower energy consumption, hence less cost and less environmentally unfriendly; they affect profit. The bulk of the operational expense in data centers is associated with the use of energy, whereby the use of energy is environmentally unfriendly and poses environmental concerns; clouds are forecasted to contribute to 5.5% of carbon emissions globally by 2025. Balancing energy efficiency and reliability will require novel optimization approaches for today's and future cloud computing systems with robust fault tolerance