=====================================================

PCU Consortium - PCU Platform

https://pcu-consortium.github.io

https://github.com/pcu-consortium/pcu

Copyright (c) 2017 The PCU Consortium

=====================================================

About PCU Platform

The PCU Consortium aims at building an Open Source, unified Machine Learning (ML) platform targeted at business applications such as ecommerce, that makes search a first-class citizen of Big Data.

License : Apache License 2.0

Requirements : Java Open JDK 10, Maven 3

Authors : Authors

PCU Package

PCU can build a packaged version (available only on Linux).

The requirements are :

• openjdk 10 (build & execute)
• apache maven 3+ (build)
• docker as sudo user (execute) (see https://docs.docker.com/install/linux/linux-postinstall/)

The packaged version contains :

• PCU platform
• PCU provided files collector agent
• PCU provided web collector agent
• PCU provided database collector agent

For each of these modules are provided scripts, configurations and sample datas. The package uses docker images of cots (elasticsearch, kafka, zookeeper, mysql) for its uses.

-pcu-entreprise-search-release/
---config/
---data/
---dist/
---lib/
---bin/

The scripts available in bin/ are :

• start-pcu-cots.sh : launch and initialize docker images of elasticsearch, zookeeper, kafka and mysql
• stop-pcu-cots.sh : stop and remove the docker images
• start-pcu-platform-server.sh : start PCU platform
• start-agent-filesystem.sh : execute PCU provided files collector agent on sample data
• start-agent-http.sh : execute PCU provided web collector agent on http://www.open-source-guide.com/Solutions website
• start-agent-database.sh : execute PCU provided database collector agent on sample mysql database

Quickstart

Clone sources

git clone git@github.com:pcu-consortium/pcu.git

Demo Package

Build

Build requirements

Build the package:

cd deployment
./make-release.sh

This creates the package folder : deployment/pcu-entreprise-search-release

Execute

Execution requirements

1. Launch the COTS :

cd deployment/pcu-entreprise-search-release/bin
./start-pcu-cots.sh

This create and start all the docker container used by the Platform and its provided agents :

• elasticsearch-pcu (docker.elastic.co/elasticsearch/elasticsearch:6.4.2)
• zookeeper (confluentinc/cp-zookeeper:5.0.1)
• kafka (confluentinc/cp-kafka:5.0.1)
• mysql-pcu (mysql:5.6)

If you have elasticsearch, zookeeper, kafka or mysql running on your OS with default ports you need to stop them.

2. Execute PCU platform server :

cd deployment/pcu-entreprise-search-release/bin
./start-pcu-platform-server.sh

This start the PCU platform server on port 8080 : http://localhost:8080

3. Execute PCU provided files collector agent :

cd deployment/pcu-entreprise-search-release/bin
./start-agent-filesystem.sh

To check if data has been collected, execute search : PCU

4. Execute PCU provided web collector agent :

cd deployment/pcu-entreprise-search-release/bin
./start-agent-http.sh

To check if data has been collected, execute search : Open

5. Execute PCU provided database collector agent :

cd deployment/pcu-entreprise-search-release/bin
./start-agent-database.sh

To check if data has been collected, execute search : ACAJOU

Build PCU

• To build PCU platform :

cd sources
mvn clean install -DskipTests

• To build provided files agent :

cd provided/agent-filesystem-norconex
mvn clean install -DskipTests

• To build provided web agent :

cd provided/agent-http-norconex
mvn clean install -DskipTests

• To build provided database agent :

cd provided/agent-database-jdbc
mvn clean install -DskipTests