Skip to main content

Self-host Windmill

Getting started​

Windmill GitHub repository contains all the necessary files to run Windmill locally, or to self-host it.

We encourage docker-based deployments, and provide a docker-compose file to help you get started.

Docker​

tip

Simplified instruction for docker-compose in the README.

info

When talking about the 'compose', or 'caddyfile', we explicitly refer to the ones on Windmill's GitHub repository.

Using docker and caddy, Windmill can be deployed using two files, (docker-compose.yml and Caddyfile) and in a single command.

Caddy takes care of managing the TLS certificate and the reverse proxy, Postgres of storage, Windmill-LSP provides editor intellisense. All managed by one docker-compose.yml file.

Configuration​

Let's assume you wish to deploy Windmill to the windmill.example.com domain.
This information only needs to be propagated to the docker-compose.yml file, using the WM_BASE_URL environment variable.

Create/edit the .env file at the root of the project and set your desired address: 

# .env
DB_PASSWORD=supersecret
WM_BASE_URL=windmill.example.com

Setting the WM_BASE_URL configures Windmill to use it as its base url, but also configures Caddy to use it as the domain.

Deployment​

Once you have setup your environment for deployment, you can run the following command:

$ docker compose up

That's it! Head over to your domain and you should be greeted with the login screen.

tip

Default credentials are [email protected] and changeme

Update​

To update to newer version of Windmill, all you have to do is run:

docker compose pull or docker compose pull windmill if you wish to update only the Windmill image.

Database volume is persistent, so updating the database image is safe too.

Helm chart​

We also provide a convenient helm chart for Kubernetes based self-hosted setup.

Detailed instructions can be found in the README file in the official repository of the chart.

tip

If you're familiar with helm and want to get started quickly, you can deploy in one command, with paths relative to the official repository:

helm install windmill-chart windmill/  \
      --values windmill/values.yaml    \
      --namespace=windmill             \
      --create-namespace

Enterprise deployment with Helm​

The Enterprise edition of Windmill uses different base images and supports additional features. One important feature is better caching for depencies in a super cache supported by S3.

You need:

  1. An Enterprise license key
  2. An AWS account and S3 bucket
  3. AWS credentials or IAM roles prepared for access from the Windmill worker pods

See the helm chart repo readme for more details. The exact setup for S3 access will vary according to your environment.

Compile from source​

  1. Go to frontend/: npm run install, npm run generate-backend-client then npm run build
  2. Install the lld linker
  3. Go to backend/: SQLX_OFFLINE=true cargo build --release
  4. The windmill binary is at: target/release/windmill

It is enough to run it with DATABASE_URL=<your_database_url> ./windmill

Windmill binary will make the assumption that nsjail is in PATH, that python3 is available at /usr/local/bin/python3 and deno at /usr/bin/deno. it will also assume that you are connected to postgres using a superuser. If you cannot use a superuser, see the following section.

Run windmill without using a postgres superuser​

Create the database with your non-super as owner:

CREATE DATABASE windmill OWNER nonsuperuser

As a superuser, create the windmill_user and windmill_admin roles with the proper privileges, using:

psql <DATABASE_URL> -f init-db-as-superuser.sql

where init-db-as-superuser.sql is this file

It will not give windmill users access to all your databases, just that one because the schema public is not shared across database, it is contained wholly inside each database: https://www.postgresql.org/docs/current/ddl-schemas.html#DDL-SCHEMAS-PUBLIC

Then last, run the following commands

GRANT windmill_admin TO nonsuperuser;
GRANT windmill_user TO nonsuperuser;