An Approach Towards Failover & Back-Up Recovery Services for Cloud Computing Environment

Abstract

Technically, Cloud is a network of Servers, offering computing services. Cloud Computing is widely defined as an Internet based Infrastructure/Platform/Software Services that are provisioned to the end users on pay per use basis. In the present scenario, continuous access to information is a must for the smooth functioning of business operations. Proper load balancing and configuration of multiple paths increases the data availability, assists in reducing unnecessary redundant replicas thereby resource consumption, implements failover, and enables scalability. This paper discusses some of the existing fail-over methodologies in cloud computing environment and presents the cost-effective method of private cloud services through efficient fail-over Manager. We performed an experiment on Private Cloud model. Our experiment consists of an application deployed in a Virtual Machine (VM) of a Server A, which is clustered with the Virtual Machine of Server B. The health of both the Servers is monitored by a Fail-over Manager module. The Fail-over Manager monitors the parameters like Availability, Reach-ability, CPU Utilization, Memory Utilization, Packet Errors and Latency and accordingly notifies the Administrator, through a Short Service Message (SMS), whose mobile number is already registered with the manager. The Batch file programmed to make a backup of the hosted application, will be triggered by the Task Scheduler. This helps the Administrator to restore the application at later point of time. This helps us to achieve efficient resource utilization and avoids un-welcome call for the Server Administrators by ensuring a fair allocation and efficient fail-over management of every computing resource.