Support Vector Machine Based Cloud Quality Of Service Violation Prediction With Rule-Based Rectification

With the advancement of technology, networking and the Internet, the Cloud computing Software as a Service (SaaS) model has changed the sales model for software providers by allowing customers to access applications with browsers and smartphones. Recently, SaaS has become the preferred delivery mode...

Full description

Saved in:
Bibliographic Details
Main Author: Wong, Tong Sheng
Format: Thesis
Published: 2019
Subjects:
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:With the advancement of technology, networking and the Internet, the Cloud computing Software as a Service (SaaS) model has changed the sales model for software providers by allowing customers to access applications with browsers and smartphones. Recently, SaaS has become the preferred delivery model and consumers are increasingly demanding more custom-built SaaS for delivering specific business outcomes. However, the dynamic nature of Cloud computing has led to Cloud services’ such as SaaS, acceptance being slowed down. The main concern of the Cloud SaaS providers is on guaranteeing and of the Cloud consumers is on receiving quality services in terms of performance such as response time, throughput and availability as agreed in the Service Level Agreement (SLA). Thus, actions have to be taken by the Cloud SaaS providers to detect, predict, prevent and rectify such Quality of Service (QoS) violations. To overcome such a challenge, many existing research works have been in progress and focused on the resources management to minimize quality of service (QoS) violation so as to maintain the level of performance and availability guaranteed in the SLA. This thesis addresses the main concern by proposing an approach in monitoring Cloud environment resources to counter measure Cloud QoS violation. This proposed approach is named as Cloud QoS Detection, Prediction, Prevention and Rectification (CQoS-DPPR). CQoS-DPPR is embedded with 16 decision rules on response time and throughput for detection and prediction of QoS violations. Based on these rules, a scaling and fault tolerant (SFT) algorithm is activated to decide whether scaling of resources is required to prevent the cloud service to be further downgraded to certainly violation condition due to resource is not appropriately provisioned.