Readme.MD
Note: PDF above gives a more conceptual overview. For some reason in github, Chrome has an issue with PDFs, Firefox seems to work fine. Just fyi.
This solution provides organizations with many pi instances to quickly enable a distributed control and programming model. The user can dynanmically instantiate extension code (in ext directory) across the network, manage each circuit without having to be physicaly near the Pi and attached circuits, and can enable the remote pis to report data to endpoints over the network which are interested in the results.
iotos can run on windows and linux, pi circuit specific functions only operate on the pi. The tests were done on windows 7 with node 6.9 above (Node 7 is best). Will widen out later.
A youtube video set is pending to setup, operate and extend.
Setup Preperation (if you have node -v >= 6.9 you can skip this)
Node is the base programming language for this solution, and you must be able to run "node -v", and verify you have version 6.9 or higher.
-
See the bringup.txt if your pi systems are not on the pixel release on the pis, suggested.
-
If on linux and have node 10.29 (node -v), need to upgrade, we can talk in class how to upgrade, net:
sudo apt-get remove nodered
sudo apt-get remove npm
sudo apt-get remove nodejs
-
Pixel and arm7 or later (Latest Pi B)
-
Visit this nodesource: https://github.com/nodesource/distributions
-
Select the Debian (Pixel is a Debian version) commands.
-
Then when done with nodejs 6.9 or 7.0 or later install, upgrade npm
- nodejs -v (or node -v) [if not in path, fix that first]
- npm install npm
-
-
arm6 (still use Pixel), the following is required
- download latest nodejs (6.91 at the time of this writing) from nodejs.org
- cd /usr/local
- sudo mkdir nodeX.YZ
- cd nodeX.YZ
- sudo tar -xvf
- cd (go to home directory where you will run node from)
- Update PATH to reference to /usr/local/nodeX.YZ/bin to .bashrc and .profile
- Debug using node -v or nodejs -v, if does not work, make it so...
- If you open a window, and "git --v" fails, you can skip to Double Check Dependencies.
- Download from git hub (install and run git on pi)
- Download git for windows (google - download windows git, install) and pi (see above if did not already, e.g. apt-get)
- Download node for windows (google - windowss 64 bit msi installer, install)
- Open new a shell window (new is important, your path will be updated to find new tools!)
Issue the commands to see if tools installed ok (your "toolchain")
- node -v (getting 6.91 or higher is good) [if not found, reinstall node]
- npm -v (getting 3.+ is good) [intalled with node]
- git --v (yes, two dashes, should 2.7.x or higher) [if not found, reinstall git]
- Create your rasperry pi circuits, the default we use for testing below:
- GPIO 24 -> led (long leg) -> led (shortleg) -> 220K ohm resister -> ground
- This is called "red_warning_light" resource for basic version, and also,
"red_warn_publish_light" same circuit but demonstrates reporting data to
cloud server as operations are performed on target circuit.
- iotos server will run on the Raspberry Pi, which will probably have breadboard
and circuits to be contorled
- You will install iotos on each pi, and share the configuration.json file across
all of them
- The iotos.js file will be started on each Pi once the confiugration.json is
established, copied to each unit and will support your specific application
- The manage.js command can be used to control from one device all Pis in the iotos
environment. manage.js can be ran on windows to control a set of pis, seems like
a popular configuration as I work with students. For windows, use git to pull down
and in the iotos directory, before issuing "npm install", update the package.son
as described below
-
For each pi or windows system involved in the solution...
-
Validate node, and other software preqreqs via Double Checking text above
-
Open a window
- Create a proejct directory in the file system, e.g. mkdir projects cd projects
-
Pull the code from github with your git command (this will create a iotos subdirectory with the code)
-
git clone https://github.com/pdykes/iotos (you may have to run twice if not ran every before)
-
cd iotos (created by git command above, fyi)
-
[on windows only - reminder, only useful for management command, server must run on Pi]
- Use notepad or another editor, edit the package.json file and remove the single "onoff" line.
-
"dependencies": {
"body-parser": "^1.15.2",
"cfenv": "^1.0.3",
"collections": "^5.0.5",
"command-line-args": "^3.0.3",
"debug": "^2.3.2",
"express": "^4.14.0",
"node-rest-client": "^2.0.1",
"onoff": "^1.1.1", <- remove this line, save the file (on winows only)
"os": "^0.1.1",
"properties-reader": "0.0.15"
- npm install (npm wil install all the dependencies
- create a cirucit on GPIO, we use a basic LED circuit on gpio port 24
- (optional) set debug on if you wish (see below)
- See start the server below, or if brave:
- nodejs iotos.js (on pi only, just starts pi circuit server,
also, could be node or nodejs, depends on install level)
This starts the server, you should see "Welcome to iotos"
- TODO command line not ready yet, install in Chrome web store the Postman application
- nodejs iotos.js (on pi only, just starts pi circuit server,
also, could be node or nodejs, depends on install level)
This starts the server, you should see "Welcome to iotos"
- To test managing the solution, see the management command examples below
- To enable debug
-
cd working directory
- background: manage.js is the CLI for managing all the pis/circuits,
- background: iotos.js is the server module to run on PI,
- background: and led_light is example IoT extension in ext\ directory
-
Windows set DEBUG=manage
-
Pi (debian) export DEBUG=manage,iotos,led_light
-
Debug lines will be colored
-
- open a window (once)
- cd projects\iotos direcotry (once, if you have not pulled the code, etc... see above)
- use ifconfig (linux), and find the ip address of the unit (lan, wireless lan, etc..)
- edit configuration.json
- Update this section for player_1_pi (assume one pi is called player_1_pi for this readme)
{
"player_1_pi": {
"resources": {
"internal_IT" : {
"ip" : "192.168.1.111", <- change your ip in this location
"port" : "8085",
"iotos_root_uri" : "/api/v01",
"secure" : "true"
- save the file...
- npm install npm (just make sure have latest npm)
- npm install
- node iotos.js (may need to use "nodejs" vs. "node")
prompt:~/projects/iotos $ nodejs iotos.js
iotos Bootstrap [os:linux arch:arm]
iotos Version v0.0.2 11/10/2016 Instance: player_1_pi
iotos Initializing Resource Mapping: configuration.json
iotos Port assigned from config file: 8085
iotos IP Address assigned from config file: 192.168.1.111
iotos Instance: player_1_pi on 8085 initializing...
iotos Ctrl-C to exit.
iotos Instance player_1_pi established on port 8085
iotos Utilize "miotos" or "nodejs miotos" to manage via command line
iotos Web access at: http://192.168.1.111:8085/api/v01/welcome
iotos Welcome!
- Congratulations, you now are able to issue management commnads.
-
On windows, only the management command line supported (can do this on linux/pi as well) for management operations
-
On raspberry pi, the server "for that node" and the management application is supported This server will run iotos and it will load modules to interact with resources and circuits
-
When running the management command line on windows, open the iotos\package.json file and remove the "onoff" package.json entry (leave all others in place!).
-
The latest release also offers a prototype target to get events that occur and simulate forwarded status to another real-time server endpoint
-
To prepare to run the test endpoint (install iotos as above on new targer server (can be windows))
- open a window on a linux or pi system
- change directory to project diretory (e.g. ..\projects for example)
- clone the source (git clone... see below)
- cd projects\iotos
- npm install (if you get an error, first try "npm install npm", secondly, if on windows, see package.json trick above, then retry this command)
- nodejs test\event_emitter_web_app.js
-
To configure and try out:
- Note the ip address of where this test endpoint is running (ipconfig, ifconfig)
- Change the project_dir\iotos\configuration.json file:
-
"restful": {
"type" : "restful",
"root_url" : "192.168.1.14:8756/gameserver",
"secure" : "false"
},
Substitute 192.168.1.14 to your ip address of the server event_emitter_web_app.js
is running on.,
- To test this out using the sample"
- You can use the same LED Circuit we use for test
- Start the iotos server on pi with led circuit
- cd projects\iotos
- (if have not installed prereqs - issue a "npm install", do only once)
- nodejs iotos.js (server should start)
- In a second window on PI or on a windows box, each of the manage commands should result in data being forwarded to the event_emitter_web_app.js example
cd projects\iotos
nodejs manage.js -t player_1_pi -r red_warn_publish_light -d init
nodejs manage.js -t player_1_pi -r red_warn_publish_light -d start
nodejs manage.js -t player_1_pi -r red_warn_publish_light -d toggle
- Output
Management Command [os:linux arch:arm]
iotos Version v0.0.2 11/10/2016 Target host: player_1_pi
iotos Initializing Resource Mapping: configuration.json
iotos Port assigned from config file: 8085
iotos IP Address assigned from config file: 192.168.1.111
iotos CLI Init request operation initiated to http://192.168.1.111:8085/api/v01/control
{
"command": "init",
"resource": "red_warning_light",
"host": "192.168.1.111:8085",
"unit": "player_1_pi",
"port": "8085",
"url": "http://192.168.1.111:8085/api/v01/control",
"active": true,
"init_message": "Resource red_warning_light Resource Basis led_light for instance 1 initialized.",
"response": {
"module": "led_light",
"method": "init",
"status": "OK",
"return_code": "200"
},
"config_message": "Resource red_warning_light Resource Basis led_light for instance 1 configured."
}
Management Command [os:linux arch:arm]
iotos Version v0.0.2 11/10/2016 Target host: player_1_pi
iotos Initializing Resource Mapping: configuration.json
iotos Port assigned from config file: 8085
iotos IP Address assigned from config file: 192.168.1.111
iotos CLI Start request operation initiated to http://192.168.1.111:8085/api/v01/control
{
"command": "start",
"resource": "red_warning_light",
"host": "192.168.1.111:8085",
"unit": "player_1_pi",
"port": "8085",
"url": "http://192.168.1.111:8085/api/v01/control",
"response": {
"module": "led_light",
"method": "start",
"status": "tbd",
"return_code": "200"
}
}
Management Command [os:linux arch:arm]
iotos Version v0.0.2 11/10/2016 Target host: player_1_pi
iotos Initializing Resource Mapping: configuration.json
iotos Port assigned from config file: 8085
iotos IP Address assigned from config file: 192.168.1.111
iotos CLI Toggle request operation initiated to http://192.168.1.111:8085/api/v01/control
{
"command": "toggle",
"resource": "red_warning_light",
"host": "192.168.1.111:8085",
"unit": "player_1_pi",
"port": "8085",
"url": "http://192.168.1.111:8085/api/v01/control",
"response": {
"module": "led_light",
"method": "toggle",
"status": "tbd",
"return_code": "200"
}
}
Management Command [os:linux arch:arm]
Command options:
--verbose/-v - Verbose Mode at startup
--directive/-d - Directive mode*
--resource/-r - Name of resource*
--port/-p - Target port override
--targethost/-t - Target host override
--configFile/-f - Configuration file
--help/-h - Help
* required
-
GPIO Library
After a pretty lenthy review, iotos utilizes Raspberry Pi GPIO library is the "on/off" package. This is for a few reasons:
- npm install is supported, allowing easy install and bringup for new folks npm install --save onoff
- supports nearly all the GPIO programming model (iotos can let you use other models as well)
- well used and many examples on the web using this library and includes several good libraries
See https://github.com/fivdi/onoff for more information on github. This knowledge will be useful for more advanced work. iotos tries to unleash the solution, not contain it. Thus, you should be able to extend iotos once you get more experienced.
New Features
0.0.4 (mvp4 01-02-2017)
- Alexa voice support (experimental)
- liveActive commmand result in management tool (lists active resources and their state)
- Tested further set intervals
Between releases next step
- Post a youtube video
- Overview
- Add support for a circuit
- Control a circuit
- Control more than one Pi
- Manage from on Pi and via a separate machine (windows box)
- Enable Alexa and control with voice (this requires an AWS account (no cost if careful))
0.0.5 (mvp5 look ahead I'm thinking)
- Angular based web management UI
- Johnny-five examples added (show can you use more than one library)
- make a windows package.json (on-off package keeps causing folks to email me)