Enhancing Concurrent User Capacity Through Kubernetes Implementation: A Case Study at PT. XYZ

This study explores deploying and evaluating a Kubernetes-based architecture at PT XYZ to enhance scalability and increase maximum concurrent user capacity. The migration involved refactoring legacy services into Docker containers and orchestrating them on a Kubernetes cluster spanning multiple worker nodes. This study replaced the VM-based environment with containerized workloads to enable dynamic resource allocation, horizontal pod autoscaling, rolling updates, and self-healing. Load tests using Apache JMeter, simulating API requesr under simulating API requests under steadily increasing user loads, and captured throughput, response times, error rates, CPU, and memory metrics via Prometheus and Grafana to measure performance gains. The Kubernetes deployment sustained up to 60 % more concurrent users than the legacy setup, with consistently lower latency, fewer errors, and more stable resource usage. These results confirm that Kubernetes offers an elastic, resilient, and efficient infrastructure, well-suited for high-demand enterprise applications requiring high availability, rapid adaptability, and decreasing operational costs.