ENHANCING NETWORK PERFORMANCE LOAD BALANCING IN CYBER CAFE NETWORKS WITH DIJKSTRA ALGORITHM ON MIKROTIK

The internet has become a fundamental necessity in various activities today. Stream Cyber Cafe, as an internet service provider, faces the challenge of maintaining network quality so that users can comfortably engage in activities such as gaming, streaming services, and social media. Load balancing, utilizing multiple internet sources from various ISPs, becomes a solution to enhance network efficiency and responsiveness. This research focuses on implementing the Dijkstra algorithm on MikroTik devices to determine the shortest path based on DNS server latency from various internet service providers (ISP). The main steps include configuring MikroTik devices, analyzing latency connections to DNS servers, and employing the Dijkstra algorithm. The Dijkstra algorithm, utilizing a Greedy approach, considers the minimum weight from the starting node to other nodes. Testing using the "PING" command provides information on the number of hops or steps required to reach each DNS server. Dijkstra adapts the shortest path based on latency, yielding optimal load balancing efficiency. Configuring MikroTik features, such as Firewall Mangle, Routing Table, and Routing Gateway, supports the functionality of the Dijkstra algorithm. Test results show that each ISP has a different shortest path to DNS servers, and the Dijkstra algorithm can determine the shortest path by considering time or latency factors. Although the author acknowledges some technical challenges during implementation, the proposed solution successfully overcomes these challenges. Thus, the use of the Dijkstra algorithm on MikroTik proves its effectiveness in enhancing the performance and reliability of the network in the Stream Cyber Cafe environment.