Development of efficient fault tolerant methods and practices forcloud computing environment

Abstract

In the current day scenario, the needs of the Internet users have dramatically changed the newlineway the Information Technology resources are used and provisioned. These ever changing newlinedemands of the Internet users have led to greater adoption of cloud computing in newlinebusinesses. Cloud computing being a new delivery model offers services on-demand pay- newlineper-use basis. newlineUsers are offered services over the network without the hassle of newlinemaintenance of the infrastructure or the platform or the software. Thus Quality of Service newline(QoS) is a very crucial factor in this model. Being a metered model, providing availability newlineand reliability of the services demand a highly fault tolerant services to be provided to the newlineclients. Fault tolerance is the capability of the system to carry on functioning even despite newlineoccurrence of failures. Achieving 100% fault tolerance is far away from reality; rather newlinefailures can be minimized without change in user experience. Fault detection, prevention newlineand recovery are the three stages in fault tolerance. Building resilience within the system newlinealso results in a fault tolerant system. Public cloud providers build and maintain large data newlinecentres and provide services over the Internet. There are many techniques and methods newlinethat address fault tolerance in cloud computing that are centred on data-centre kind of newlineinfrastructure. In our work, faults in cloud computing model are studied and analysed. We newlinehave proposed fault detection and prevention policies and methods for a private cloud newlinemodel using commodity hardware. newlineThe commodity hardware requires a different newlineapproach to handle faults since the capacity and design of these systems are different from newlinethe specialized data centre kind of infrastructure. An attempt has been made to reuse the newlinelow-end commodity hardware within the organizations and provide fault tolerant newlineinfrastructure services. A novel approach that lifts the cloud from the datacentre to low- newlineend hardware has been attempted in this work. Keeping the Stack lightweight is one of newlinethe key criteria.