The scalability metric, based on cost-effectiveness in distributed systems using fundamental laws
No Thumbnail Available
Date
2018
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
KNUST
Abstract
Today’s computer systems are more complex, more rapidly evolving, and more essential to the conduct of business than those of recent past. The complexity becomes more rigid in the case of distributed systems. A distributed system is a collection of independent computers that appear to its users as a single coherent system. A distributed system should be deployable in a wide range of scales, in terms of numbers of users and services, quantities of data stored and manipulated, rates of processing, numbers of nodes, geographical coverage, and sizes of networks and storage devices. The derived scalability metric of this thesis is based on cost effectiveness, in which the effectiveness is a function of the system's throughput and its Quality of Service. It is a strategy based scalability metric that generalizes the well -known metrics for scalability of parallel computations to describe heterogeneous distributed systems. Scalability is measured by the range of scale factors that gives a satisfactory value of the metric, since a good scalability is a joint property of the initial design and the scaling strategy. What makes this derived metric unique is the fact that, it separates the impact of throughput and response time on the metric, formalizing the notation of a scaling strategy, introducing QoS evaluation and more also, introducing formal scalability enablers which are optimized at each scale factor. Throughput curves for all systems with bottleneck demand 𝐷𝑚𝑎𝑥 are constrained to lie below the line 1𝐷𝑚𝑎𝑥⁄. If one needs to improve the performance further than this limit, then it is necessary to reduce the demand at the bottleneck center somehow.
Description
A thesis submitted to the department of Computer Science Kwame Nkrumah University of Science and Technology in partial fulfilment of the requirements for the degree of
Master of Philosophy in Computer Science