CD Skripsi
Perbandingan Algoritma Weighted Least Connection Dan Weighted Round Robin Pada Load Balancing Berbasis Docker Swarm Di Google Cloud Platform
ABSTRACT
Technological advancements have altered Indonesians’ television viewing habits from analogue to digital television and the popularity of video-on-demand services. However, the enhancing use of video-on-demand services poses a challenge to sustaining servers’ performance in the face of high demand. The solution is implementing efficient load balancing and scheduling algorithms, such as Docker Swarm, as a container virtualisation technology. This study examined three video resolution categories, 240p, 720p, and 1080p, and requests ranging from 1 to
10.000. The study’s findings demonstrated that the higher the video resolution, the larger the video file size, which reduced throughput and enhanced response time, request failure, and CPU utilization. Based on the specified parameters, the analysis revealed that the weighted least connection algorithm is preferable to the weighted round-robin algorithm. Implementing a docker swarm positively affected server performance maintenance compared to a single server. This study presented further insight concerning implementing technology and algorithms to enhance video-on- demand servers’ performance and ensure users' optimal viewing experience.
Keywords - load balancing, docker swarm, weighted least connection, weighted round robin, video on demand
Tidak tersedia versi lain