Integrating CloudFS with Zovida

Set up

System Requirements

Java 8 or higher
Maven 3.5+
Neo4J Should already be setup

The LMS app directory looks something like the tree below:

|-- Procfile
|-- package.json
|-- pom.xml
|-- b2intgr
|   |-- README.md
|   |-- bin
|   |-- pom.xml
|   `-- src
|-- app.js
|-- .env
|-- app.json
|-- jsdoc-conf.json
|-- node_modules
|-- scripts
|-- src
|-- test
`-- webpack.config.js

There are a few new files:

b2intgr/
- Contains the BackBlaze integration
pom.xml
- This is only here to let Heroku know that maven & java are required for this app
- It is defined as the parent POM to b2intrgr/pom.xml and tricks Heroku into compiling the true target
Procfile
- This is now Heroku's starting point. It starts both servers after building:

    web: b2intgr/bin/cloudfs.sh start && npm start

To run heroku local, modify the .env file to set the following variables. Eg:

GRAPHENEDB_BOLT_PASSWORD=reggae
GRAPHENEDB_BOLT_USER=neo4j
GRAPHENEDB_BOLT_URL=bolt://localhost:7476
PORT=5555

The PORT value is for expressJS. Both CloudFS and expressJS use the GRAPHENE settings.

Alternatively, one can modify the config files:

src/main/resources/config.json - (Main config)
src/test/resources/config.json - (Test config)

(The test config contains instructions and comments)

Deploy to Heroku

Assumptions:

directory initialized as git repo:

    git init`

A Heroku app exists or was created:

    heroku create

Assuming nodejs buildpack already added:

    heroku buildpacks:add heroku/nodejs

Add the Java buildpack:

    heroku buildpacks:add heroku/java

The java buildpack should come after heroku/node

Confirm:

   heroku buildpacks

Output:

=== rdnpm Buildpack URLs
1. heroku/nodejs
2. heroku/java

Add graphenedb:

    heroku addons:create graphenedb:dev-free

Finally:

    git push heroku master

Confirm it's working:

https://rdnpmjava.herokuapp.com/

The /buckets data is pulled via CloudFS

https://rdnpmjava.herokuapp.com/buckets

Development and Testing

Set properties in the config.json file. $VARIABLES may be used directly in the config file. They will be interpolated from corresponding environment variables. That is, for BASH:

    export VARN=value

The config ..:

    "poolSize": 5,
    "maxPoolSize": $VARN,
    "neo4jConf": {
        "urlString": "$GRAPHENEDB_BOLT_URL",
        "username": "$GRAPHENEDB_BOLT_USER",
        "..."
    },

... Same as setting the values in .env.

Dev cycle

During Dev it might be easier to keep CloudFS running in the background.:

## Start:

    b2intgr/bin/cloudfs.sh start

## Stop:

    b2intgr/bin/cloudfs.sh stop
    
## Test:

    b2intgr/bin/cloudfs.sh test
    
## Clean build

    b2intgr/bin/cloudfs.sh clean

Start expressJS as needed: npm start

archimedon / npmjava