Aloes - Device manager

# Aloes - Device Manager

pipeline status coverage report

  • Replicate devices with rich semantic properties
  • Authentification by access tokens and API keys
  • Handle devices and sensors CRUD methods triggered via HTTP and MQTT
  • MQTT events generated by HTTP operations hooks
  • Dispatch MQTT payload from several protocol sources
  • Adding context to sensors with OMA schemas
  • Store devices and sensors state ( MongoDB )
  • Fast access to sensor resources, clients and timers ( Redis )
  • Mesurement (sensor values and tags ) stored in timeseries ( InfluxDB )
  • Users and Devices files stored in filesystem or cloud providers, see loopback-component-storage
  • Trigger persisted schedulers (via sensors) to create timed scenarios ( Skyring )
  • Interact with external application and share selection of devices

Swagger Explorer

Full Docs

Application build upon :

# Prerequisites

These need to be installed, configured and started manually, when working locally and not using docker setup.

  • Install MongoDB to persist Users, AccessTokens, Devices, Sensors, Addresses, File models.
mongod --config /usr/local/etc/mongod.conf
  • Install Redis to persist SensorResources, Schedulers and Client models, MQTT connection and RateLimiter
redis-server /usr/local/etc/redis.conf

influxd -config /usr/local/etc/influxdb.conf
  • Install Skyring to use persistent timers for Scheduler
# start NATS server
nats-server

# start Skyring
DEBUG=skyring:* skyring run -p 3005 -s localhost:3455

# Folder structure

  • /. --> Main application configuration, dependencies list, and launch scripts

  • /bin --> contains some maintenance scripts - not needed yet

  • /deploy --> contains environment variables ( hidden files )

  • /log --> contains logs from PM2 and Nginx used via Docker

  • /storage --> contains folders where Users / Devices files are stored

  • /src --> contains source code

    • /. --> Loopback configuration
    • /boot --> code executed at application start
    • /initial-data --> JSON datasets to make global application running
    • /lib --> Helpers for models and services
    • /middleware --> scripts used in development/staging only
    • /mixins --> add special properties to models
    • /models --> Data models, controllers and REST API descriptions
    • /services --> external modules
    • /views --> templates used for automated mailing

# API

# HTTP

URL pattern : +apiRoot/+modelPluralName/+path

Access controlled using Access tokens ( set in HTTP headers ) and roles for Users and using apiKey property for Applications and Devices.

# MQTT

Topic pattern : +userId/+modelName/+method/[+modelId]

Access controlled using Access tokens ( set in MQTT password ) and roles for Users and using apiKey property for Applications and Devices.

# Configuration

Edit your config in .env_sample and save it as .env. You can override these by populating deploy with files corresponding to an environment ( eg: .env_production ... ).

# Installation

  $ npm install

# Linting

  $ npm run lint

# Debug

  • To set Aloes verbosity, configure SERVER_LOGGER_LEVEL from 0 to 4

  • To activate Loopback debug :

  $ DEBUG=loopback npm run start:dev

More info...

# Starting project

Create or update .env file to match your enviroment.

For example to run in local mode, create ./deploy/.env_local using .env_sample as an example.

# With Nodemon

  $ npm run start:local

# With PM2

  • Starting PM2 processes :
  $ npm run start:pm2
  • Stopping PM2 processes :
  $ npm run stop:pm2

# With Docker

You can serve the project with a distant server by filling deploy folder with files corresponding to an environment ( eg: .env_docker ), and then launching this app with docker via docker-compose up

Remember to update *.dockerfile to match your environment, if you don't use docker-compose, otherwise edit docker-compose.yml to match your environment and needs.

Creating environment :

  $  npm run build:docker

Starting containers :

  $  npm run start:docker

Stopping containers :

  $  npm run stop:docker

# To deploy with your own TLS / SSL certificates

Read https://nodejs.org/api/tls.html#tls_tls_ssl

openssl req -x509 -nodes -days 365 -newkey rsa:2048 -keyout ./deploy/key.pem -out ./deploy/cert.pem

openssl dhparam -out ./deploy/dhparam.pem 2048

# To deploy with Localtunnel

Install your own server instance : https://github.com/localtunnel/server

And then install wildcards certificates with : https://certbot.eff.org/

Finally, configure TUNNEL_HOST and TUNNEL_SECURE in your environment files.

# TODOs

  • Add e2e tests for processes interruption signal and tests for MQTT with mysensors / lorawan source

  • Finish restore helper for InfluxDB

  • Implement user+ip rate limit for HTTP endpoints requiring auth

  • Finish account linking with github

  • Add user(s) in a team to easily share devices access ( via collaborators property/role ? )

  • Catch and store data related to MQTT traffic ( https://github.com/mcollina/aedes-logging#readme ; via specific sensor instance ? )

  • Update to loopback 4

  • Add redis replication config ( docker deployement )

    • https://stackoverflow.com/questions/45902031/docker-swarm-redis-and-sentinel-with-master-slave-replication-ip-resolution-cl

    • https://redis.io/topics/replication#configuring-replication-in-docker-and-nat

  • Synchronize data from Aloes ecosystem to a distributed ledger ( IOTA Tangle ? )

API →