blyzer/compor-academy
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: integrate Bunny Stream course videos
All checks were successful
CI / Validate frontend and API (push) Successful in 1m50s
Details
CI / Build and publish Docker images (push) Successful in 18s
Details

Browse Source
This commit is contained in:
Cauê Faleiros
2026-09-02 10:09:36 -03:00
parent 3d4c72e85c
commit 2c137529bc
15 changed files with 607 additions and 40 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
PORTAINER.md
View File

@@ -2,6 +2,30 @@
Use `docker-compose.yml` as a Portainer Stack from this repository. It deploys the frontend, API, and PostgreSQL as one internal Docker network. Only the web container exposes a port; it proxies `/api` to the API container.
## Bunny Stream video hosting
To let instructors upload protected course videos, configure these API environment variables in the Portainer Stack. They are server secrets: never add a `VITE_` version or place them in the frontend container.
```env
BUNNY_STREAM_LIBRARY_ID=123456
BUNNY_STREAM_API_KEY=your-bunny-library-api-key
BUNNY_EMBED_TOKEN_KEY=your-bunny-embed-view-token-key
BUNNY_MAX_UPLOAD_MB=5120
```
In Bunny Stream, create one Video Library using the Volume tier. In that library's security settings, enable MediaCage Basic DRM and Embed View Token Authentication, then copy its token key into `BUNNY_EMBED_TOKEN_KEY`. Disable Direct Play and MP4 fallback if they are not needed, and add the Academy production hostname to Bunny's allowed referrers. This keeps playback inside Bunny's protected iframe and makes the Academy API issue a short-lived embed token only after a logged-in user requests a lesson.
The deployed Nginx proxy permits uploads up to **5 GB**. Keep `BUNNY_MAX_UPLOAD_MB` at or below `5120`; reduce it if you want a smaller application-level limit. Large uploads stream through the API rather than being held in memory.
After deploying the new images:
1. Sign in as an instructor and open **Gerenciar Cursos**.
2. Enter the lesson title first, then choose **Selecionar vídeo** in the Bunny Stream section.
3. Wait for the upload message, add the lesson, and save the course. Bunny encoding continues asynchronously.
4. The editor refreshes Bunny processing status automatically; it also has an **Atualizar** action for a pending lesson.
The API key and embed-token key never reach the browser. The instructor browser uploads to the authenticated Academy API, which sends the file to Bunny; learners receive only a signed iframe URL.
This is a Docker Swarm stack: Portainer pulls prebuilt API and web images from the Gitea Container Registry. It never builds Dockerfiles itself.
## Gitea Actions registry secrets