<![CDATA[Online marketplaces have become popular among the Indonesian public lately. This trend has an impact on the increasing visitor traffic, making the server's performance in handling requests heavier. Another impact that arises is the decline in web server performance accompanied by lag or network disruptions that cause web page responses to become slow. This issue can be addressed using load balancing, which is a method that distributes traffic across multiple servers to prevent load accumulation on a single server. Testing 10,000 requests with Apache Bench resulted in a response time of 7.725 ms for a server cluster (3 servers) using the round-robin algorithm. In comparison, the least connection algorithm produced a response time of 7.496 ms, while a single server resulted in 119.656 ms. In a stress test using JMeter with 10,000 requests, the server cluster produced a response time of 60 ms, while a single server resulted in 39.418 ms. In the failover implementation, load balancing works well in distributing requests to the available servers, ensuring that even during high traffic or server failures, the system maintains optimal performance and minimizes response times. Failover is the process of redirecting requests from a dead/broken server to another active server.]]>
Copyrights © 2026
