Infrastructure

TL;DR -Everything runs on a single rented server (a VPS) behind nginx, with the app split into a handful of Docker containers -backend, admin, docs, simulator -plus a PostgreSQL database. This page is the entry point; the child pages carry the full detail.

Who this is for -DevOps & Operations: your starting point. Executive & Product: a plain picture of what we run and how simple it is. Engineering: the map to environments and provisioning.

Social Live's entire backend lives on one server -a VPS, which just means a computer we rent by the month from a hosting company (currently Hetzner). There is no sprawling cloud setup: one box does the work.

That one box is deliberately kept simple in two ways:

• Containers. The app is broken into a few containers. A container is a sealed, self-contained package that holds a program plus everything it needs to run -think of a shipping container that you can lift onto any ship without repacking. We use Docker to run them. The pieces are the backend (the API and live-video engine), the admin dashboard, this docs site, and a simulator used for QA. Alongside them runs a PostgreSQL database that stores all the real data.

• A reverse proxy. In front of all those containers sits nginx, acting as a reverse proxy. A reverse proxy is the single front door: every visitor arrives at nginx first, it handles the secure-connection (HTTPS) part, and then it quietly forwards each request to the right container in the back. Visitors never touch the containers directly.

The one exception to "everything goes through nginx" is live video and audio, which travels over its own direct network path for speed -that detail lives in the Architecture topology.

Where the detail lives

Two child pages carry the full technical picture, and neighboring sections cover the rest:

Related pages

• Environments
• Provisioning
• Operations
• Architecture