Analysis of the Performance of a Cloud Computing Processing Queue with Correlated Reneging of Tasks and Resubmission

The Quality of Service (QoS) experienced by the packets that constitute a cloud computing task does not only depend on the performance of the cloud computing network infrastructure at the data centre but also on the performance of the access networks and the internet core networks traversed by the packets before the task is fully submitted to the cloud computing or processing servers for execution or processing. Sometimes, some of the tasks that are submitted and are waiting in queues to be processed could be removed from the queues without being processed (task reneging or dropping). Removing tasks from the queue could result from the user's impatience, missing execution deadline, security reasons, or an active queue management strategy. The reneged or dropped tasks could be resubmitted depending on the reason for which it was dropped. The resubmission of tasks decreases QoS and increases the energy consumption overhead as more energy is required to transport the resubmitted tasks through the access, core, and data centre networks. In this paper, we propose a multiserver queueing model with correlated reneging and resubmission to analyse the performance of a cloud computing server.