Traffic Classification for the Dispatcher in a Server Farm based on SVM

摘要: The increased capacity of Internet connections leads to more and wider usage services. An effective task-server assignment policy is obligatory improve the performance server farms. Scalable techniques that categories traffic always assist enhance performance. Existing dispatcher farm do not consider a timely classification approach for improvement QoS. So this paper proposes classifier which classifies requests using dynamic properties farm. proposed meachine learning method based on SVM aims minimize makespan execution fairness. performed in terms precision waiting time are measured simulated results have shown significant improvement.