Doctor of Philosophy
Electrical and Computer Engineering
With the transformation of cloud computing technologies from an attractive trend to a business reality, the need is more pressing than ever for efficient cloud service management tools and techniques. As cloud technologies continue to mature, the service model, resource allocation methodologies, energy efficiency models and general service management schemes are not yet saturated. The burden of making this all tick perfectly falls on cloud providers. Surely, economy of scale revenues and leveraging existing infrastructure and giant workforce are there as positives, but it is far from straightforward operation from that point. Performance and service delivery will still depend on the providers’ algorithms and policies which affect all operational areas.
With that in mind, this thesis tackles a set of the more critical challenges faced by cloud providers with the purpose of enhancing cloud service performance and saving on providers’ cost. This is done by exploring innovative resource allocation techniques and developing novel tools and methodologies in the context of cloud resource management, power efficiency, high availability and solution evaluation.
Optimal and suboptimal solutions to the resource allocation problem in cloud data centers from both the computational and the network sides are proposed. Next, a deep dive into the energy efficiency challenge in cloud data centers is presented. Consolidation-based and non-consolidation-based solutions containing a novel dynamic virtual machine idleness prediction technique are proposed and evaluated. An investigation of the problem of simulating cloud environments follows. Available simulation solutions are comprehensively evaluated and a novel design framework for cloud simulators covering multiple variations of the problem is presented. Moreover, the challenge of evaluating cloud resource management solutions performance in terms of high availability is addressed. An extensive framework is introduced to design high availability-aware cloud simulators and a prominent cloud simulator (GreenCloud) is extended to implement it. Finally, real cloud application scenarios evaluation is demonstrated using the new tool.
The primary argument made in this thesis is that the proposed resource allocation and simulation techniques can serve as basis for effective solutions that mitigate performance and cost challenges faced by cloud providers pertaining to resource utilization, energy efficiency, and client satisfaction.
Abusharkh, Mohamed, "Climbing Up Cloud Nine: Performance Enhancement Techniques for Cloud Computing Environments" (2016). Electronic Thesis and Dissertation Repository. 3881.