Abstract
A cloud System is considered as one of most required shared distributed system in which multiple clients are connected with multiple servers in an integrated environment. The intermediate layer is formed between clients and server to arrange the order of their execution. The Server side is defined with specific cloud servers and integrated virtual machines along with memory and IO capabilities. Multiple client requests are generated under the request time parameters. As a user enters to the system, a particular VM is allocated to it for the process execution. It the VM is not capable to execute the process under defined constraints, the migration of the request is performed. In cloud system, one of the critical issues is to handle these multiple request in effective way so that the overall wait time will be reduced. In this present work, multiple queue system is suggested to reduce the wait time and to reduce the chances of request migration. The presented work is divided in three main stages. In first stage, the Virtual machines are set in order of their capacity parameters. Once the machines the ordered, in next stage, the user requests are set in specific queue based on dead line criticality. In this work, two queue middle layers are composed under dead line criticality analysis called hard deadline queue and soft deadline queue. Once the processes are allocated to these queues, the next work is to perform the allocation of the processes to virtual machines. This allocation is performed under capacity and load parameter. After this allocation, the execution of the processes on particular virtual machine is done under process time and deadline parameters. The processes that are not been executed before the deadline are considered as migrated processes. The obtained result from system shows the effective execution of processes on cloud system along with reduced deadline and process migration probability.