Node-RED Appsody Stack

This is a stack for building a deployable Node-RED application using Appsody.

The resulting application will run the Node-RED flows in a headless mode without the editor enabled.

The stack builds into the application a standard set of cloud-native endpoints to determine readiness and liveness of the application. These are exposed as:

Readiness endpoint: http://localhost:3000/ready
Liveness endpoint: http://localhost:3000/live
Health check endpoint: http://localhost:3000/health

Templates

This stack provides a single template - simple. It is the most minimal template to get started.

Getting Started

1. Build the stack image

In the root directory of this repository run:

sh scripts/buildStackImage.sh .

That will create the docker image for the stack.

2. Package the templates

In the root directory of this repository run:

sh scripts/packageTemplate.sh .

That will create a tgz file for each template in the stack - currently only the one. The files are creates in the repo directory.

3. Register the stack with the appsody cli

This repository includes an index.yaml file that can be registered with the appsody cli.

Before you register it, you must edit it to update the urls property to point at where the template tgz file is on your local file system.

Once you have updated that, you can register the stack by running the command (with the right url for your local file system):

appsody repo add local file:///Users/nol/code/knolleary/appsody-node-red-stack/repo/index.yaml

Using the stack

The stack is designed to augment a Node-RED project.

A typical Node-RED project will have the structure:

├── README.md
├── flow.json
├── flow_cred.json
├── package.json
└-- settings.js

The package.json file will have the structure:

{
    "name": "AppsodyProject",
    "description": "A Node-RED Project",
    "version": "0.0.1",
    "dependencies": {},
    "node-red": {
        "settings": {
            "flowFile": "flow.json",
            "credentialsFile": "flow_cred.json"
        }
    }
}

This stack will use the package.json file to install any additional dependencies the project has and also to identify the flow file - using the node-red section.

If the package.json file doesn't have node-red section, the stack will use flow.json as the default flow file.

The settings.js file is a normal Node-RED settings file, but the stack will override some of the values:

httpAdminRoot is set to /admin - this moves the editor away from /. In the future the default behaviour will be to disable the editor entirely for production environments.
more tbd

Given this structure, the follow steps can be used to use the stack:

Initialise it using the Appsody CLI:
```
appsody init node-red
```
Once it has been initialised, you can then run the application using:
```
appsody run
```
This launches a Docker container, exposing it on port 3000.

Next steps

This is a very early stage POC of a Node-RED stack. These's a tonne more work to do to take full advantage of the appsody development process - such as live reloading of the application.

We also need to identify what are the right set of default settings the stack should be applying.

knolleary / appsody-node-red-stack