Performance analysis: An algorithm for computer performance evaluation using QN model

Computer performance is determined by the amount of useful work accomplished by system or network with the time and resources used. A general methodology is proposed to describe how service demands are calculated from different types of data using queuing network (QN) model using different configuration. A model is constructed to answer several scenarios, like the addition of multiple disks and multiple processors. The proposed methodology, identify which configuration i.e. Disks and/or processors, processes the data in the least amount of time. The response time of different configurations are calculated and also obtain least residence time. These results are obtained by different performance parameters. The computation of utilization depicts how fast the system is being used, lesser the computation, greater the response time. The right configuration is found out which gives the least response time for the given data. This is the configuration which maximizes the full potential of the system and satisfies the end user. We developed a tool which helps the developer during the development phase of any project to identify the ideal system configuration.