الفهرس 
abstractsOnly 14 pages are availabe for public view
 |  | from | 106 |  |  |
| | | from | 106 | | |
Abstract
Service Level Agreements(SLAs) more and more prove their added value in the modern distributed computing including cloud computing specifically. Indeed, well-established systems for managing SLAs represent a challenging aspect in cloud computing. Certainly, designing such systems will save many costs when being spent in paying SLA violation penalties. Additionally, these systems will decrease the traditional SLA failures and thus ensure the level of quality of offered cloud services. In order to extract a managed SLA, it should be handled through accomplishing several stages (i.e. life cycle) which are established as: Definition and Awareness, Negotiation, Design, Implementation, Monitoring and Compliance. In this thesis, an engineering model with five structured phases for managing SLA in a cloud environment is introduced. This SLA engineering model will benefit both cloud providers and consumers. It will allow providers to structure and organize the SLA correctly in an attractive way for customers. Simultaneously, it will provide customers with the necessary and sufficient awareness to get a fair SLA for resources or services they may access. Also, a case study is presented to demonstrate how the proposed five phases utilized to construct the required SLA in a private cloud. Further, a reliable service-based architecture for managing the SLA life cycle in cloud computing is proposed. The architecture is composed of five main layers which are defined as: Infrastructure Layer, SLA Layer, Provider Layer, Consumer Layer, and Security and Privacy Layer; each layer in turn consists of multiple components provide certain management functionalities (e.g. violation detection and recovery). In this architecture, a particular skeleton for each component of the suggested SLA layer is depicted. In particular, a proposed monitoring system is detailed constructed and implemented to observe the SLA performance and detect any violations by determining a specific calculated compliance value. Additionally, it can diagnose and recover such violations by taking the proper actions.