aa94a8eebd
Docker changes: - Use oven/bun:1-debian as base image instead of node:20-trixie-slim - Replace npm ci with bun install --frozen-lockfile - Replace npm run commands with bun run - Mount bun.lock instead of package-lock.json - Update cache paths for Bun's cache directory - Rename NPM_RUN_PRUNE env var to BUN_RUN_PRUNE CI workflow changes: - Update bashlib.sh npm-install function to use bun - Update superset-frontend.yml to use bun run commands - Update release.yml to use setup-bun action and changesets - Update superset-e2e.yml to use setup-bun action - Update superset-playwright.yml to use setup-bun action - Update superset-translations.yml to use setup-bun action Note: superset-embedded-sdk and superset-websocket remain on npm as they are separate packages with their own lockfiles. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Getting Started with Superset using Docker
Docker is an easy way to get started with Superset.
Prerequisites
Configuration
The /app/pythonpath folder is mounted from ./docker/pythonpath_dev
which contains a base configuration ./docker/pythonpath_dev/superset_config.py
intended for use with local development.
Local overrides
Environment Variables
To override environment variables locally, create a ./docker/.env-local file (git-ignored). This file will be loaded after .env and can override any settings.
Python Configuration
In order to override configuration settings locally, simply make a copy of ./docker/pythonpath_dev/superset_config_local.example
into ./docker/pythonpath_dev/superset_config_docker.py (git-ignored) and fill in your overrides.
WebSocket Configuration
To customize the WebSocket server configuration, create ./docker/superset-websocket/config.json (git-ignored) based on ./docker/superset-websocket/config.example.json.
Then update the superset-websocket.volumes config to mount it.
Docker Compose Overrides
For advanced Docker Compose customization, create a docker-compose-override.yml file (git-ignored) to override or extend services without modifying the main compose file.
Local packages
If you want to add Python packages in order to test things like databases locally, you can simply add a local requirements.txt (./docker/requirements-local.txt)
and rebuild your Docker stack.
Steps:
- Create
./docker/requirements-local.txt - Add your new packages
- Rebuild docker compose
docker compose down -vdocker compose up
Initializing Database
The database will initialize itself upon startup via the init container (superset-init). This may take a minute.
Normal Operation
To run the container, simply run: docker compose up
After waiting several minutes for Superset initialization to finish, you can open a browser and view http://localhost:8088
to start your journey.
Running Multiple Instances
If you need to run multiple Superset instances simultaneously (e.g., different branches or clones), use the make targets which automatically find available ports:
make up
This automatically:
- Generates a unique project name from your directory
- Finds available ports (incrementing from defaults if in use)
- Displays the assigned URLs before starting
Available commands (run from repo root):
| Command | Description |
|---|---|
make up |
Start services (foreground) |
make up-detached |
Start services (background) |
make down |
Stop all services |
make ps |
Show running containers |
make logs |
Follow container logs |
make nuke |
Stop, remove volumes & local images |
From a subdirectory, use: make -C $(git rev-parse --show-toplevel) up
Important: Always use these commands instead of plain docker compose down, which won't know the correct project name.
Developing
While running, the container server will reload on modification of the Superset Python and JavaScript source code. Don't forget to reload the page to take the new frontend into account though.
Production
It is possible to run Superset in non-development mode by using docker-compose-non-dev.yml. This file excludes the volumes needed for development.
Resource Constraints
If you are attempting to build on macOS and it exits with 137 you need to increase your Docker resources. See instructions here (search for memory)