Management of Service Level Agreements for Big Data Analytics Applications in Cloud: A Layer-based Study
Date
2019
Authors
Zeng, Xuezhi
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Nowadays, deployment and operation of big data analytics applications (BDAAs) in cloud are increasingly becoming a trending practice. These applications offer organizations the capabilities of constructing valuable information and extracting actionable insight for enhancing the evidence-based decision-making process. Many leading providers such as Google and Amazon provision such analytics capabilities in the form of service to customers in a pay-per-use economic model. In today's competitive world, the potential business values of these applications depend a lot on the quality of service offered by providers. Hence, to gain competitive advantages, providers need to be more customer-focused and proactive in their marketing strategies not only to create customers awareness of their services but also meet customers' best expectations for service quality. That is to say, providers must provide the required and promised services, and the services must satisfy users' requirements, such as availability, scalability, elasticity, and so on.
Given these circumstances, it is very important and necessary for efficient methods to manage and guarantee the quality of service promised. Service Level Agreement (SLA) that represents the contract between providers and customers which captures the agreed upon guarantees regarding the quality of service is one of the effective methods. SLAs play an integral role in governing the relationships between providers and customers in the context of these applications in cloud. Besides setting the expectations by dictating the quality and the type of service, SLAs are also increasingly considered as a strong differentiator allowing a provider to offer different levels of service guarantees and to differentiate itself from its competitors. Therefore, how to manage SLAs for cloud-hosted big data analytics application (BDAAs) in ensuring SLA guarantee has become a crucial and essential aspect.
However, the management of SLAs for cloud-hosted BDAAs is extremely challenging due to the increased complexities and uncertainties imposed by the applications. For instance, the applications usually span heterogeneous and distributed software frameworks across multiple layers, which considerably impacts the allocation and configuration of datacenter resources in order to accommodate changes in the big data workloads and to guarantee analytic results within SLA constraints.
Most of the extant studies focus on SLAs management in traditional distributed computing environments like grid computing or cloud computing, which is not the case for SLA management for BDAAs in cloud. Although the research on the management of SLAs for BDAAs in cloud is now attracting growing attention, to the best of our knowledge, the study on SLA management for BDAAs in cloud is still in its infancy.
In this thesis, we focus on the research problem that how to manage SLAs for big data analytics applications in cloud in ensuring SLA guarantee. This research problem has been break down into five research questions that have been respectively addressed in Chapter 3 to Chapter 7. Accordingly, we have made five major contributions including one systematic literature review contribution, one conceptual contribution and three technical contributions.
To the best of our knowledge, this thesis is one of the first attempts to systematically study SLA management for big data analytics applications in cloud. It is believed that the outcomes of this research will yield very positive contribution in terms of technical content, significance, and impact to the advancement of scientific research in this field.
Description
Keywords
Citation
Collections
Source
Type
Thesis (PhD)
Book Title
Entity type
Access Statement
License Rights
Restricted until
Downloads
File
Description
Thesis Material