Developing a Secure and Accessible Local Online Coding Environment

Abstract - This research refers to an implementation of a high performance, scalable and secure online judge system for automated code executing and programming exams. Our principal goal is to provide a highly configurable environment which can accept any amount of concurrent submissions while ensuring the strictest isolation from other submissions and outside interference. For this study, we utilized a microservices architecture coordinated by Kubernetes which is supported by dockerized runtime environments and utilises Judge0 as the code evaluation engine. We used PostgreSQL coupled with Prisma ORM for efficient data handling, and made use of advanced containment to remove vulnerabilities. As discussed in the results, we created a system that provides high performance through scalable deployments, and highly configurable security against known vulnerabilities. This study serves to provide all users with a flexible framework that provides a resilient, highly efficient and secure online judge system capable of supporting educational programming exams and competitive programming. Key Words: Online Judge, Kubernetes, Docker, Code Execution, Security, Scalability, Microservices