Skip to content
Trailer.devDocumentation

Search is only available in production builds. Try building and previewing the site to test it out locally.

Standalone Mode

A freshly started standalone instance with its built-in local host already connected.

Standalone mode runs the server and agent in a single process. It is published as a container image:

Terminal window
docker pull ghcr.io/trailer-dev/standalone:latest

The ghcr.io/trailer-dev registry path is set at release time. Your deployment’s registry and organization may differ. Adjust the image reference to match where your images are published.

This is the simplest way to run Trailer.dev on a single machine: one process is both the control plane and the host that runs workspaces.

The standalone process starts the server first, waits until it is ready, then starts the agent. Both run for the lifetime of the process.

sequenceDiagram
    participant P as Standalone process
    participant Srv as Server
    participant Ag as Agent
    P->>Srv: start server
    Srv-->>P: ready
    P->>Ag: start agent
    Ag->>Srv: connect and reconcile

Standalone shares the same configuration file (default ~/.config/trailer.dev/config.yml) and flags as the separate modes. The most relevant flags:

FlagDefaultPurpose
--server-url, -Shttp://localhost:8090Server bind address. The agent connects to the same URL.
--data-directory, -dtrailer_dataServer database directory.
--build-directory, -b/tmpDirectory used when building images.

There is no environment-variable binding for these flags. The server-only TRAILER_* settings (admin bootstrap, SMTP, backups, file storage) still apply here. See Environment variables.

Minimal config.yml:

server_url: http://0.0.0.0:8090
data_directory: trailer_data

Expose the port and mount the Docker socket so the built-in agent can manage containers. Add --gpus all for NVIDIA GPU passthrough:

Terminal window
docker run -d \
-p 8090:8090 \
-v /var/run/docker.sock:/var/run/docker.sock \
--gpus all \
ghcr.io/trailer-dev/standalone:latest
  • The container needs Docker socket access to run workspaces.
  • Persist the data directory across restarts.
  • For multi-host setups, run a dedicated Server and one or more separate Agents instead.