In this tutorial we will show how you can run the Spark RDF Analyzer right from eclipse using a Docker container for the Tomcat webservice.

• Download Docker from https://www.docker.com/products/docker and follow their installation instructions.
• Get an official Tomcat image (we will use Tomcat 8.0 with Java 8):

docker pull tomcat:8.0-jre8

• Create a file /my/path/to/tomcat-users.xml, which will be injected into the Docker container to allow us to access the Tomcat manager and executing scripts using the text API of Tomcat.

<?xml version="1.0" encoding="UTF-8"?>
<tomcat-users xmlns="http:https://tomcat.apache.org/xml"
              xmlns:xsi="http:https://www.w3.org/2001/XMLSchema-instance"
              xsi:schemaLocation="http:https://tomcat.apache.org/xml tomcat-users.xsd"
              version="1.0">
	<user username="manager" password="manager" roles="manager-gui" />
	<user username="demo" password="demo" roles="manager-script" />
</tomcat-users>

• Create one folder /my/path/to/data where you put in all the datasets that should be available to the running RDF Analyzer.
  The directory structure on you local system should look similar to the following:
  

/
+-- my
|   +-- path
|   |   +-- to
|   |   |   +-- tomcat-users.xml
|   |   |   +-- data
|   |   |   |   +-- sib200
|   |   |   |   |   +-- sib200.nt
|   |   |   |   +-- sib400
|   |   |   |   |   +-- sib400.nt
|   |   |   |   +-- storage
|   |   |   |   |   +-- parquet

• By executing the following command, you can run the Tomcat container.
  This will overwrite the default Tomcat user config and mount the local directory /my/path/to/data into the container, accessible at /home/data.

docker run \
	-dit \
	--name tomcat \
	-p 8080:8080 \
	-v /my/path/to/tomcat-users.xml:/usr/local/tomcat/conf/tomcat-users.xml \
	-v /my/path/to/data:/home/data \
	tomcat:8.0-jre8 \
&& docker logs -f tomcat

• You will have to setup a local maven profile that will be used to overwrite the maven variables during the deployment process. This way we can ensure that real credentials are hidden from public and excluded from the POM.
  • Open your local settings file, located at ${user.home}/.m2/settings.xml.
    If the file is not already there, simply create it!
  • Insert the following lines:

<?xml version="1.0" encoding="UTF-8"?>
<settings>
        <servers>
                <server>
                        <id>tomcat-localhost</id>
                        <username>demo</username>
                        <password>demo</password>
                </server>
        </servers>

        <profiles>
                <profile>
                        <id>tomcat-localhost</id>
                        <properties>
                                <tomcat.deploy.server>tomcat-localhost</tomcat.deploy.server>
                                <tomcat.deploy.url>http:https://127.0.0.1:8080/manager/text</tomcat.deploy.url>
                        </properties>
                </profile>
        </profiles>
</settings>

• Check out the project and import it into eclipse as a maven project.
• Set up a new run configuration for it.
  • Right click on the project > "Run as" > "Maven build..."
  • Set following properties:
    • Name: RDF Analyzer (Tomcat)
    • Goals: clean tomcat7:redeploy
    • Profiles: tomcat-localhost
    • Parameter: p.type=war
• Click "Run" and check the console output in eclipse. The war file gets deployed to the running Tomcat instance.
• Go back to the console, where Tomcat logs to and wait for the completion of deploament.
• Open your web browser at http:https://127.0.0.1:8080/spark-rdfanalyzer2/ and you should see the RDF Analyzer running. Unless you changed the mount path, your datasets will be available under /home/data inside the container.
  By clicking on "Add new Graph", you will be prompted for the path where your data files reside. Enter /home/data/sib200 to add the graph for the sib 200 dataset.