Node-RED Guide
Turn an architecture sentence into a clean flow shape you can read at a glance.
Node-RED — start here
Node-RED is a visual programming platform for integration and logic: you build by wiring pre-built nodes into flows. The connectors and the syntax are handled for you, so your effort goes into what the system should do — the logic — not the plumbing to connect things or the boilerplate of a language.
Beginnings
Every way work enters a piece, each as its own path.
Flow shape →
Shared services (link call)
Shared pools and calls, invoked once, results return.
Flow shape →
Subflow vs link call
When a link call is enough, and when a subflow earns its keep.
Flow shape →
Single sink
The one sink each family of paths converges on — it only sends, never routes.
Flow shape →
Find the seams
Find the three or four components hiding in a blob.
Design patterns →
Levels of reuse
Pick the lowest rung of extraction that solves it.
Design patterns →
Decouple frontend & backend
Treat the dashboard-to-logic boundary like a client and a server.
Design patterns →
Use the context store
Hold the object in context instead of threading it through wires.
Design patterns →
Config: env vars vs persisted context
Env vars for deploy-time, persisted context for user edits.
Design patterns →
Generating flows (Admin API)
Emit flow JSON and deploy it through the Admin API.
Design patterns →
Error handling
A Catch node per tab turns a thrown error into a route you control.
Operating flows →
Backpressure & memory
Pace the fast side to the slow side before the heap blows.
Operating flows →
Deploy modes
Pick the smallest restart that ships the change.
Operating flows →
Debugging
Tap the wire, surface node state, keep the flow observable.
Operating flows →