Designing LMS Model Based on Serverless to Solve Performance Issue on High Concurrency Workload

Abstract

Moodle is one of the Learning Management Systems (LMS) that play a critical role in supporting online education. One of the important factors in Moodle LMS is its performance in high concurrency workload environments. High concurrency workload refers to situations where the number of users exceeds the system's capacity. This study aims to propose an architectural model for Moodle LMS that can address performance issues under such heavy-load conditions. The proposed approach adopts a hybrid architecture that integrates serverless technologies for database, file storage, and session handler components, along with containerization on virtual machines (VM) using IaaS for the core system, supported by load balancing and autoscaling mechanisms. The research follows four stages: problem identification, system design, implementation, and LMS model analysis. Evaluation results show that the developed LMS model can handle up to 1500 concurrent users without significant performance degradation, maintaining a response time below 2500 ms and a failure request rate below 1%. Further testing with minimum resource configuration allows the system to support up to 10.000 concurrent users. Benchmark scores using the Moodle plugin indicate optimal performance across all aspects. This model has proven to enhance the reliability and scalability of LMS platforms in high-concurrency workload environments.