EVALUASI KINERJA LOAD BALANCING DENGAN ALGORITMA SCHEDULLING NEVER QUEUE

Load balancing is used as a technique to handle large loads that cannot be carried out by a single server, so that the server does not experience overload. In handling load sharing, Load balancing uses a scheduling algorithm (Scheduling). The scheduling algorithm that is generally used is Round Robin which works by dividing requests evenly and then creating a queue for the server so that unfinished processes wait in the queue for quite a long time. In the Load balancing system there is an algorithm that adopts two speed models, which works by looking at the server status and the smallest connection delay, namely the Never Queue Algorithm. This study aims to determine the performance of Load balancing when the Schedulling Never Queue Algorithm is applied, based on predetermined scenarios and parameters. This study succeeded in implementing the Never Queue Algorithm in a Load balancing system for the Apache web server where the Time Per Request value will be lower if the Request received is larger when compared to using the Round Robin Algorithm. The Request Per Second value increases when the Requests sent are getting bigger. In terms of sharing server connections, the load balancer will share the load on the number of Requests based on the Shortest Expected Delay (SED) Algorithm, so several web servers receive different numbers of connections, so that processes don't stay in queues for a long time.