High System Availability, Reliability, and Security in Cloud Environment

Abstract

In IT industry, service provisioning is an emerging domain; in this context, cloud provisions must always be acquirable and procurable 24x7 over the internet with easy and secure access to cloud users. During accessing the cloud services, the uninterrupted availability is a significant concern for all service brokers and some of the services like recovery point, scheduling, load balancing, replicationetc, are also applied to utilize performance. Thus, besides these services, there is prospect to research middleware and infrastructure or arbitrator solutions. The primary purposes of this research presented in this paperis to develop a model-based algorithm and apalatable system approach which provides high availability for cloud resources and services. The proposed availability-based model is calculated based on cloud resource uptime and downtime. The availability of cloud is based on the actual running time of datacenter and if any fault in a server is identified, load need to migrate from faulty server to another server immediately. During this migration process, unavailability that may transpire in a server considered as downtime of that server. The goal of using Weibull distribution in this research simulation is for reliability calculation and data analysis due to its versatile accuracy. The Exponential distribution purposely applied to model the behavior of units that possess constant failure.