Managing Request Congestion in a Multi-Channel Queuing System through a Hybrid Service Approach
Abstract
This article proposes a new mechanism that combines FIFO and priority-based service methods to improve service efficiency in remote web server service systems. The proposed approach assumes that requests arriving at the system are divided into priority and non-priority groups, and the order of servicing requests is assumed to be FIFO when service channels are not fully occupied, and priority when they are fully occupied. To accurately represent the flow of requests and service processes, system state probabilities are developed based on Kolmogorov equations. Also, based on these equations, the general functional relationship of the system is formed using the generating function. For a public service system of the M/M/m/r type, the generating function is presented and the functional relationship for the generating function is obtained.