Abstract :
Website-based application developers usually install applications that are made on the server. Increasing requests for client information services to many servers will result in the workload on the server. Excessive workload on the server can reduce server performance in providing information services to clients. Server Clustering is a technology that can reduce workload, it takes a minimum of two servers to achieve high resource availability. Container orchestrator is the process of managing or scheduling the work of each container for applications in multiple clusters. The largest container orchestratror platforms that are widely used are kubernetes and docker swarm. Based on previous research, the high availability analysis used the Kubernetes orchestrator. This research method used the Docker Swarm orchestrator as a comparison. The parameters measured are high availability, CPU utilization, throughput, self-healing. A comparison is needed to compare the high availability given by the two orchestrators using the same load on the server. Keywords : clustering, container, docker swarm, high availability, orchestrator