Skip to main content
The drim CLI runs Docker Compose commands in your deployment directory. You can use it with an existing stack; you do not need to run drim setup again.

​
Adopt your existing Compose file

  1. Install drim and back up your database, files, and configuration.
  2. Stay in the current deployment directory. Moving directories can change the Compose project name and cause Docker to create new volumes.
  3. Ensure the main Compose file is named docker-compose.yml. drim 0.1.15 checks for that filename. If yours is compose.yml, rename it in the same directory and update any scripts that refer to it. Avoid keeping two competing default Compose files.
  4. Keep the existing .env, AUTH_SECRET, database credentials, volumes, networks, and reverse proxy configuration.
  5. Validate the configuration with docker compose config --quiet.
If your normal deployment uses COMPOSE_FILE, COMPOSE_PROJECT_NAME, profiles, or extra environment files, keep the same Compose context when running drim. For complex setups, direct Docker Compose commands may be clearer. Run these commands from that directory when needed:
Use drim restart --recreate after an environment change. A plain restart retains the container’s old environment. drim upgrade pulls the images selected in your Compose file and applies them; it does not rewrite the file or change a pinned version for you.

​
Replace an older two-container layout

If your Compose file still contains separate usekaneo/api and usekaneo/web images, follow Move to the bundled container. drim upgrade alone does not combine those services. For a Caddy service on the same Docker network, the application upstream becomes:
Keep Caddy’s existing certificate volumes. If you replaced a bind-mounted Caddyfile, recreate the proxy container so it sees the current file:
Use your actual proxy service name. A host-based proxy uses the published application port instead of Docker’s kaneo hostname.

​
Check the result

Check docker compose ps and request /api/health through your public URL. Sign in, open existing tasks, and check an existing attachment. Confirm that the database is the original one before inviting anyone to use the instance. Keep the backup and original configuration until the deployment is verified. If a newer image migrated the database, recovery requires its matching backup, not just an older Compose file.