-
Dependencies
-
Database - MongoDB
-
Build The Project
-
Connect API
-
Configure Collectors
maven, mongodb 3.0 or above, java 1.8 recommended
Hygieia requires MongoDB 3.0 or above. Starting from version 3.2 MongoDB only supports 64-bit binary which the current Raspberry PI does not support. So it's essential to download the 32 bit MongoDB 3.0 version for Hygieia. The official MongoDB 3.0 32-bit binary for Linux could not be executed by Raspberry PI. So we have to compile binaries ourselves.
Luckily we can find ready-to-use binaries here:
https://andyfelong.com/2017/08/mongodb-3-0-14-for-raspbian-stretch/
https://andyfelong.com/2016/01/mongodb-3-0-9-binaries-for-raspberry-pi-2-jessie/
Download both the core and tool binaries and copy them into /usr/bin by sudo cp [your binary directory]/* /usr/bin/.
For maven, download maven from its website and follow the tutorial here to add the path:
http://agilerule.blogspot.com/2016/06/how-to-install-maven-on-raspberry-pi.html
When you turn the PI off, make sure to turn off the server by running db.shutdownServer() in the Mongo Shell. If it ever happens that the collector api is running but not showing up in UI, it means that the database is corrupt. One solution is to go to MongoDB's default location and delete everything. Start MongoDB setup all over again.
For node, follow the following command to install the new version
pi@raspberrypi:~ $ sudo su -
root@raspberrypi:~ # apt-get remove nodered -y
root@raspberrypi:~ # apt-get remove nodejs nodejs-legacy -y
root@raspberrypi:~ # apt-get remove npm -y
root@raspberrypi:~ # curl -sL https://deb.nodesource.com/setup_8.x | sudo bash -
root@raspberrypi:~ # apt-get install nodejs -y
check the node and npm version by running node -v and npm -v
After success, run sudo npm install bower -g and sudo npm install gulp -g
Create /data/db directory in the root folder.
Start a mongoDB session by running mongod if there isn't one already running.
If mongod is not working, try sudo mongod or mongod --repair && mongod depending on the error.
Run mongo to start a mongo shell.
Create a new database with custom name, for example "database", by running use database
Create a new database user with custom name and password by running
db.createUser( { user: "user", pwd: "password", roles: [ {role: "readWrite", db: "database"} ] } )
You can view all the databases and users of each by running show dbs and db.getUsers() respectively.
Go to https://github.com/capitalone/Hygieia and get master either by git clone or download.
Cd in into this folder and run mvn clean install package or
mvn clean install package -Dmaven.test.skip=true -Dpmd.failOnViolation=false
After successfully building the project, go into /UI and run gulp serve, you should be able to see the ui opening in a browser.
mvn clean install package may fail in the UI step on the npm install --ignore scripts and npm install command, we'll have to do everything after this manually.
If it fails on a collector, go to pom.xml and skip that step, and run mvn again.
When it fails on the UI step, go to /UI and delete /node and /node_modules
run npm install and npm install --ignore scripts
After success, run npm install bower and npm install gulp
After success, run bower install and gulp build
gulp build might show error, but as long as gulp serve brings the dashboard browser up we are fine.
Go to /api and create a dashboard.properties with content
dbname=database
dbusername=user
dbpassword=password
dbhost=localhost
dbport=27017
The dbname, username, password need to match your custom database information in mongoDB.
Run this command in root directory.
java -jar api/target/api.jar --spring.config.location=api/dashboard.properties -Djasypt.encryptor.password=hygieiasecret
The port should be available. If shows address already in use error, kill the application that is occuping the port.
Run gulp serve in /UI and you should see the UI with api connected. Sign up for an account and log in to configure your dashboard.
Go to http://capitalone.github.io/Hygieia/collectors.html#tool-collectors and find the collector you want to use.
Configure the application.properties file according to the sample properties provided there.
Run the following command in the current collector's folder
java -jar target/[collectorname.jar] --spring.config.name=git --spring.config.location=../../../api/src/main/resources/application.properties
Refresh your dashboard and now you can configure the collector. (The collector might take several minutes to be enabled, but it shouldn't be longer than that.)
(Take into account that you might need to configure the api daily fetch limit of the source from which you are collecting, i.e. github repo or bitbucket repo.)
If this happen, check the port and see if any application occupies the 8080 port.
Clear the browser cache.
Jenkins default port is 8000. Should not change it.
The url you put in the ui should be the git clone path. There is a problem with api commit timezone, it might not work properly.
For the .properties set up, write git.api=/2.0/repositories/ instead of the documentation. The url you put in the UI should be https://api.bitbucket.org/{username}/{reponame}
There might be bugs in the collector codes. Collector links to jira successfully on the API layer, but UI is not displaying the content properly.