Indian Journal of Science and Technology

Abstract

Background/Objectives: Considering the growingdemandfor cloud services for development and deploying of critical business applications, it is extremely important that cloud provider guarantees a reliable and robust service by providing fault tolerance mechanisms that enable seamless execution of the business transaction execution even in presence of faulty components. The objective of this paper is to propose a collaborative fault tolerant mechanism between cloud provider and cloud client. Methods/Statistical Analysis: The collaborative fault tolerance approach considers collaboration between the cloud provider and the cloud client to develop a comprehensive fault tolerance solution that can be customised to suit to the hosted cloud applications needs.The proposed design is based on usage ofPersistent Map based strategy. Findings: The Persistent Map based strategy saves the state information of execution in the form of P-maps. The P-map is a persistent hash map that stores the current state of execution of a given task. In the case of failure, it can be used to restart the process from the last state at which the task failed and resume the application execution from that point as though no failure occurred.The P-map storage is a crucial element to be considered in the design of the system, that requires careful analysis and can have a huge impact on the execution of an application. Application/Improvements: The authors have considered an approach which requires a collaboration between cloud providers and cloud client to design a fault tolerance mechanism that takes into consideration the complex cloud infrastructure as well behaviour and functionality of the application in focus
Keywords: Cloud Computing, Fault Tolerance, Persistent Maps, Recovery System