Containers

tug.sh supports multiple ways to run containers, from quick image-based runs to custom project setups.

Option 1: Deploy from image

Use this flow when you already have a ready image in a registry.

1. Open Apps or deployment flow.
2. Choose Deploy from Image.
3. Provide the image reference (for example ghcr.io/org/service:1.2.3).
4. Set container name, ports, environment variables, and restart policy.
5. Deploy.

The Agent pulls the image on the selected VPS and starts the container with your configuration.

Option 2: Manual project/container setup

Use this when you need custom runtime control.

Typical flow:

1. Create or select a project on a VPS.
2. Define your container configuration (ports, env, resources, networks, restart policy).
3. Deploy and verify logs/status in the container view.

For advanced cases, edit compose-based configuration in the project/container configuration views before redeploying.

Option 3: Marketplace templates

Use Marketplace for common services (for example databases, cache, or ingress tools) with minimal input fields.

Templates are useful for quick bootstrapping and reproducible defaults.

Manage existing production containers

If your VPS already runs containers, tug.sh can discover them through the Agent snapshot.

This allows progressive adoption:

• connect existing VPS,
• inspect running containers,
• organize into projects,
• then standardize configuration over time.

Day-2 operations

From container views you can:

• start/stop/restart/remove,
• inspect logs and runtime status,
• edit environment variables,
• update restart and resource policies,
• configure TugRouter domains where applicable.