Resource Allocation to Maintain the Load Balance Using Scheduling in Cloud Computing

Abstract

There are an increasing number of Cloud Services available in the Internet. Cloud services can be a component of a system and different Cloud Servers that would provide different services. In this present work we have defined a multiple cloud environment. Each cloud server is defined with certain limits in terms of memory and the CPU specifications. Now as the users enter to the system, the user request is performed in terms of processes. To represent the parallel user requests, n number of requests are been generated by the users. All these requests are to be handled by the cloud servers in parallel by using the multiple cloud concepts. A middle layer is defined between the cloud servers and the client requests that will perform the allocation of the processes to different clouds in under load and over load conditions. As user requests are performed, some parameters are also defined with each request. These parameters are the process time, deadline, input output specifications etc. In the general case, the allocations of the processes are performed in a sequential order. Each process must be executed within the deadline limit. But if more than one processes occur at same time and not get executed before the deadline, in such case the processes is switched from one cloud server to other called the process migration. In this present work, a parametric analysis is performed to identify the requirement of process migration and based on this analysis the migration will be performed on these processes. The effectiveness of the work is identified in terms of successful execution of the processes within the time limits.