Analisis Kinerja Load Balancing pada Server dalam Mengoptimalkan Performa Website

Abstract

The objective of this study is to evaluate the impact of load balancing implementation on website service performance to enhance system efficiency and reliability. The research object is a website accessed through seven service endpoints: Login, Homepage, Blog, About, Shop, Contact, and Post Checkout, with a total of 3,500 requests. The testing was conducted under two scenarios: a single server architecture and an architecture utilizing a load balancer (HAProxy) implementing the Round Robin algorithm. The testing method employed Apache JMeter to simulate load and measure various performance parameters, including average response time, throughput, standard deviation, bandwidth usage, average response size, and error rate. The results indicate that the use of a load balancer significantly reduced the average response time from 7,551 ms to 3,278 ms, increased throughput from 12.5 to 27.3 requests per second, and improved response time stability by reducing standard deviation across all endpoints. Bandwidth usage also increased without changes in average response size, with the error rate remaining at 0% in both scenarios. The study concludes that implementing load balancing with HAProxy effectively enhances website service performance in terms of efficiency, capacity, and system stability, and is recommended as a performance improvement solution in resource-constrained environments.