ozwillo-ozenergy

OzEnergy - energy consumption monitoring for consumers, providers and territories alike

Note:in the following explanations, we suppose that you are using a GNU/Linux debian-derived distribution like Ubuntu on your computer. If you are using Microsoft Windows, please consider running this software in a Virtual Machine.

How to use it

Through a docker image

If you have Docker installed (if not, you can find how to do so here), and don't want to mess with your pre-existing dev environment, you may prefer this to the next section ("Or build and run it locally").

git clone https://github.com/occiware/occiware-ozwillo.git
cd occiware-ozwillo/docker/ozwillo-ozenergy
sudo docker build -t ozenergy .
sudo docker run -p 8080:8080 ozenergy

Or build and run it locally

Prerequisites

Building requires Java 8, Maven 3, Node 6 (tested with v6.2.0 and v6.3.0), Scala 2.10 or later , sbt 0.13 and Spark 1.6.1 (pre-built for Hadoop 2.6).

Running requires also MongoDB 2.6.

In the end what you need to do is :

• A priori, it is not necessary to install Scala separately, since sbt can install the most appropriate version itself. So go on the sbt 0.13 link and follow the installation instructions.

• Java, Maven and MongoDB are to be installed through command line

sudo apt-get install default-jre default-jdk maven mongodb

• After downloading the Spark 1.6.1 (pre-built for Hadoop 2.6) archive, you can extract it in your home directory or anywhere, for that matter, but don't forget to add the $SPARK_HOME environment variable by adding at the end of your ~/.bashrc file : "export SPARK_HOME=", for example : "export SPARK_HOME=/home/myself/Bureau/Demo/Softs/spark-1.6.1-bin-hadoop2.6".

• Node and npm are managed by the maven build process : you do not need to install it yourself.

Installation

• Before building the project, you need to clone and install the Spring integration package :

git clone https://github.com/ozwillo/ozwillo-java-spring-integration.git
cd ozwillo-java-spring-integration
git checkout ozwillo-java-spring-integration-1.24
./gradlew install

• Build oz-energy-aggregations subproject :

cd oz-energy-aggregations
sbt publish

It will create a fat jar using sbt-assembly and publish it into your maven-repository.

• Go in the oz-energy subproject and copy the application.model.yml file into application.yml.

cd oz-energy
cp config/application.model.yml config/application.yml

Then adapt it if you feel the need to, but it should be fine as is if you have set the SPARK_HOME environment variable as told above.

To put it in production, also set : (noauth)devmode props to false, kernel.base_uri and your application credentials (client_*) as appropriate for the chosen target Ozwillo environment, and application.url.

• Then do (in the oz-energy subproject) :

mvn clean package

Running Oz'Energy

To have the project display any data, you must have at least one running instance of the Datacore with its mongo cluster. For that, you can simply use pre-built docker images (make sure you are not currently using port 30000 and 8088, otherwise, it's not gonna work) :

# Download the Docker images
sudo docker pull smilelab/ozwillo-mongo-master-with-energy-data:1.0
sudo docker pull smilelab/ozwillo-mongo-slave:1.0
sudo docker pull smilelab/ozwillo-datacore:1.0

# Run them
sudo docker run -d -p 30000:27017 --name ozwillo-mongo-1 --hostname ozwillo-mongo-1 smilelab/ozwillo-mongo-master-with-energy-data:1.0
sudo docker run -d --name ozwillo-mongo-2 --hostname ozwillo-mongo-2 smilelab/ozwillo-mongo-slave:1.0
sudo docker run -d --name ozwillo-mongo-3 --hostname ozwillo-mongo-3 smilelab/ozwillo-mongo-slave:1.0
sudo docker run -d -p 8088:8088 --name ozwillo-datacore-1 --hostname ozwillo-datacore-1 smilelab/ozwillo-datacore:1.0

Note: Since the docker containers know each other through harcoded references to their IPs on the default docker0 network, if you already have other containers running, the actual IPs will change, and you might need to update their /etc/hosts file by hand, to do so, get a bash in the container by doing 'sudo docker exec -it /bin/bash', and edit /etc/hosts by hand.

Run Spring Boot :

mvn spring-boot:run -DrunAggregation

Note: You can remove the last parameter to avoid running Spark aggregations on startup in addition to nightly.

Open http://localhost:8080/ with your favorite browser and you're set !

If devmode has been kept to true, default sample data will be displayed for users if they don't have their own energy consumption contract defined in the Datacore yet.