Academun is a large-scale LMS and video streaming platform built with NestJS, Next.js, PostgreSQL, and HLS-based media delivery.

The project focused on rescuing and stabilizing an existing production environment affected by server instability, deployment issues, slow video processing, HLS path mismatches, infrastructure configuration problems, and security gaps.

The primary engineering challenge was recovering the platform without compromising existing media, database records, or active production services.

The Challenge

Video-heavy LMS platforms introduce additional infrastructure complexity compared with standard web applications.

The application backend, frontend, media services, processing workers, PostgreSQL database, HLS files, FFmpeg pipeline, reverse proxy, and production server must operate as a coordinated system.

An unsafe deployment or server rebuild can make existing course videos inaccessible, break database media references, disrupt HLS playback, or introduce production downtime.

Academun required a controlled recovery strategy capable of rebuilding infrastructure, protecting existing media, correcting application data, improving video processing, and restoring reliable production services.

The Engineering Approach

The recovery process began with an audit of the existing server architecture, Docker configuration, media storage, deployment workflows, and application services.

The infrastructure was rebuilt on a fresh Linux environment with Docker, Docker Compose, PM2, and Nginx.

NestJS backend services, the Next.js frontend, media APIs, and processing workers were deployed as separate application components and connected to the existing PostgreSQL infrastructure.

The recovery strategy prioritized media protection, controlled database changes, infrastructure security, and production verification throughout the deployment process.

Production Server Recovery

The existing infrastructure required a controlled rebuild to restore application stability.

The recovery process included:

• Server environment auditing
• Fresh Linux server configuration
• Docker installation and configuration
• Docker Compose setup
• PM2 process management
• Nginx reverse proxy configuration
• Frontend deployment
• Backend deployment
• Media API deployment
• Processing worker deployment
• Remote PostgreSQL connectivity
• Production service verification

The infrastructure was structured to provide clearer separation between application components and improve deployment management.

FFmpeg Video Processing Optimization

The original video processing pipeline required processing times close to the duration of the source video.

The FFmpeg workflow was optimized by reducing dependency on remote storage during processing, introducing local buffering, and tuning CPU execution.

This improved processing efficiency and created a more reliable workflow for handling video conversion and HLS generation.

HLS Video Delivery Recovery

HLS media delivery required careful handling because existing video files and database records needed to remain synchronized.

The recovery process included:

• HLS storage validation
• Read-only safety checks
• Media path auditing
• Database reference correction
• HLS path mismatch fixes
• File and database record cross-checking
• Playback workflow verification

These changes restored consistency between stored media, application records, and video delivery paths.

Media & Database Synchronization

Existing database records contained video and storage references that required correction during the infrastructure recovery.

Media files were cross-checked against PostgreSQL records before updating affected paths.

This process reduced the risk of breaking existing course content while migrating and restoring application services.

Backend & Frontend Deployment

The platform architecture includes NestJS backend services and a Next.js frontend.

The deployment process restored communication between application services, media infrastructure, processing workers, and the remote PostgreSQL database.

Production health checks were performed to verify API availability and application service status after deployment.

Docker & Nginx Infrastructure

Docker and Docker Compose were configured to manage application services and deployment environments.

Nginx was configured as the production reverse proxy for routing traffic between the frontend, backend APIs, and supporting services.

The deployment architecture provided a more structured foundation for managing the platform's production components.

Infrastructure Security Hardening

Security improvements were applied as part of the production recovery process.

The work included:

• Credential rotation
• SSH key updates
• Firewall configuration
• Restricted server access
• Database access controls
• Vulnerable dependency updates
• Production configuration review

These changes addressed security gaps discovered during the infrastructure audit and server recovery.

Production Health Verification

After deployment, application services and infrastructure components were verified through production health checks.

The verification process covered API availability, backend services, frontend deployment, media infrastructure, database connectivity, and processing workers.

Technology Stack

NestJS, Node.js, TypeScript, PostgreSQL, Next.js, Docker, Docker Compose, Nginx, PM2, FFmpeg, HLS, Linux VPS

Outcome

Academun's production infrastructure was stabilized through a controlled server rebuild, application redeployment, FFmpeg pipeline optimization, HLS path recovery, database media corrections, and infrastructure security hardening.

The resulting architecture restored reliable backend and frontend deployment, improved the video processing workflow, corrected media delivery issues, and provided a more stable infrastructure foundation for the LMS platform.