Running on Sevalla

Sevalla is a developer-focused PaaS that offers one-click templates, managed databases, and global deployments powered by GKE and Cloudflare, without dealing with complex infrastructure.

Sevalla supports Git-based deployments, Dockerfiles, container images, and most popular frameworks out of the box. This guide walks you through deploying Umami on Sevalla using their one-click template or by deploying manually with your own GitHub repository.

To get started, create a Sevalla account, and you’ll receive $50 in free usage credits to try things out.

Method 1: One-click deployment (recommended)

Sevalla provides an official Umami template that provisions an Umami app and a PostgreSQL database with a single click.

Click the button above to get started easily. Your app will be live within a few minutes. Then click the Visit App button and log in using the default credentials.

Method 2: Manual deployment

If you prefer more control, you can manually deploy from your own GitHub repo or from the public Umami repo.

Set up Sevalla project

1. Go to Applications > New Application
2. Choose GitHub repo (if you’ve forked Umami), or Public repo and paste: https://github.com/umami-software/umami
3. Set the branch to master, enter an app name, choose a region (same as your database), and select a pod size based on your needs.
4. Click Create, but skip the deploy step for now, so it does not fail since we have not added the database.

Database and deploy

1. Go to Databases > Add database
2. Choose PostgreSQL (recommended), select the same region as your app, pick a size, and click Create.
3. Once created, scroll to Connected Applications, click Add Connection, select your Umami app, and enable “Add environment variables”.
4. Set the variable name as DATABASE_URL (It may be prefilled as DB_URL, change it to DATABASE_URL)
5. Click Add connection. This links your app and DB with the correct environment variable.
6. Go back to your application and click Deploy. Once the build is complete, visit your app’s URL and log in using the default credentials.