Building A Scalable And Secure Jitsi Infrastructure With Multi-Shard Architecture For Ailleron

The Challenge

Ailleron’s existing video conferencing system faced significant performance bottlenecks, inconsistent recording capabilities, and lacked the scalability required to handle fluctuating user demand. They needed a fully branded Jitsi infrastructure optimized for Google Cloud, capable of supporting high concurrency through a multi-shard architecture. Additionally, seamless recording integration, cloud storage synchronization, and automated resource management were essential to improve operational efficiency.

Our Approach

To address these challenges, we implemented a modular and scalable Jitsi solution focused on performance, automation, and security. The latest version of Jitsi was deployed and fully rebranded to align with Ailleron’s corporate identity, ensuring a consistent user experience.

We implemented a multi-shard architecture on Google Cloud Platform using multiple Jitsi Video Bridges (JVBs). Auto-scaling mechanisms based on CPU utilization dynamically provision and decommission instances, optimizing resource usage and maintaining performance under varying loads.

For recording capabilities, we deployed five virtual machines, each hosting ten Docker-based Jibri containers. This setup enabled parallel recording streams and ensured high availability. Recordings were automatically synchronized with Google Cloud Storage for reliable and scalable archival.

To enhance connectivity and security, a TURN server was configured to support users on restrictive networks, while JWT token authentication secured meeting access against unauthorized users.

Infrastructure provisioning was automated using Terraform, enabling consistent and repeatable deployments of the entire multi-shard environment. Additionally, custom REST APIs were developed to manage Jibri servers dynamically, including starting and stopping recording instances, tracking active servers, and integrating with Ailleron’s live banking application for load-based scaling.

A webhook-based recording metadata system was also implemented, allowing clients to receive recording URLs and metadata automatically. Furthermore, cloud logging and alerting mechanisms were configured to monitor system health, ensuring high uptime and rapid issue resolution.

• Multi-shard Jitsi deployment with auto-scaling JVBs on GCP
• Docker-based Jibri scaling enabling parallel recordings
• Automated recording storage sync with Google Cloud Storage
• TURN server setup for improved network compatibility
• JWT-based authentication for secure meeting access
• Infrastructure as Code using Terraform for consistent deployments
• REST APIs for dynamic Jibri server management
• Webhook integration for recording metadata delivery
• Cloud logging and alerting for proactive monitoring

The Result

The implemented solution provided Ailleron with a highly scalable, secure, and fully branded video conferencing platform. The multi-shard architecture combined with automated Jibri scaling ensured smooth performance even during peak usage. Cloud-integrated recording, API-driven management, and robust monitoring significantly improved operational efficiency and user experience. This future-ready infrastructure positioned Ailleron for sustained growth in digital collaboration.