Please use this identifier to cite or link to this item:
|A Probabilistic Approach for Fault Analysis in Cloud Environment
|Cloud Computing;Fault Tolerance;Fault Prediction;Reliability;csed
|Cloud Computing has emerged as a revolutionary paradigm in information and communication technology. Prominence of this technology can be validated easily with its phenomenal features such as pricing-per-use, scalability and on demand availability of computing resource. But at the same time reliability and security are some of the current issues in this technology. Fault tolerance is an important approach to overcome these issues to improve overall performance of cloud computing services. Fault tolerance is the ability of a system that ensures continuity of operations even in the presence of faults in its components to increase the reliability of the computing system. Virtualization of the instances in the cloud services is considered to provide computing instances anywhere according to the users demands. There can be some fault prone instances which are needed to be handled to minimize the fault occurrences and their adverse effects on the system. Fault handling approaches include prediction of fault occurrences in the system and their prevention to make sure fault-free execution of the computing tasks in the cloud services. The proposed work presents a statistical analysis of virtual machines for finding fault occurrence probability to improve reliability. The status of the virtual machines is monitored periodically and faulty scenarios are understood with the monitored information to find criticality status of the machines. Statistical analysis for virtual machine is done based on a probability distribution model. An application interface is designed which is capable of finding availability, predicting fault occurrence probabilities and estimating sustainability of a virtual machine.
|Appears in Collections:
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.