Open source · self-hosted · MIT licensed

Your servers.
A platform to run them.

RailDock turns Linux servers into one application control plane: Git and image deploys, project networking, databases, recovery workflows, manifests, teams, and an API.

installcurl -sSL https://raw.githubusercontent.com/mona-chen/raildock/main/install.sh | bash
production / project topology
RailDock project topology showing applications, PostgreSQL, Redis and private service links
project graphhealthy · private links visible
One operational pathFrom source to recovery.
SourceGit / image
BuildDocker / builders
Runtimeservices / links
Datadatastores
Recoverbackup / PITR
Project as a system

See relationships, not a list of resources.

Applications, workers, databases, caches, volumes and private links share a project boundary. The graph is useful because the lifecycle underneath it uses the same model.

Private topologyServices can reference each other without exposing every dependency publicly.
Context stays attachedDeployments, logs, variables, networking and metrics stay with the selected service.
Data is first-classDatastores live beside the applications that depend on them, including recovery workflows.
commerce / production
applicationwebgit · :3000
applicationapigit · web + worker
processworkerqueue consumer
databasepostgresprivate · volume
cacheredisprivate
project model · apps + data + links + lifecycle
Desired state

Declare the target. Inspect the drift.

The manifest editor gives operators a project-level declaration, preview and drift surface. RailDock can compare that desired state with the running project before applying changes in dependency order.

RailDock-native manifestsapp.json compatibilityRailway-compatible formats
RailDock manifest editor with drift status and manifest quick reference
Drift detectedcompare
Preview planordered
Apply changesobserve
Operate in context

Zoom into the job at hand.

The product moves from project topology to the detail an operator needs: service lifecycle, database inspection, or a recovery trail. Each view keeps the resource and its history connected.

RailDock deployment controls and revision historyservice / deployments
01

Deploy without leaving the service.

Lifecycle actions, process scale and revision history sit together, so a deploy is not detached from the runtime it changes.

Deploy · restart · rebuild · shelllifecycle
Process counts + revision historyruntime
Rails APIcontrol plane
Jobs + realtimeorchestration
Linux hostsSSH boundary
Dokku + Dockerexecution
Own the host boundary

A control plane above your servers—not instead of them.

RailDock coordinates lifecycle work from Rails and the API, then executes on Linux hosts over SSH through Dokku and host engines. Compute location, host hardening and network policy remain yours.

ControlRails · API · jobs
ExecutionSSH · Dokku · Docker
SurfaceReact · realtime

Open where the trust boundary matters.

The control plane is MIT licensed and the API, docs and runtime architecture are inspectable.

LicenseMITpublic source
BackendRailsAPI + jobs
FrontendReactcontrol plane
AutomationREST APIScalar reference
Start from a server you control

Turn a Linux host into an application platform.

Install RailDock, connect a host, and build the project from the same control plane you will use to operate it.