dnet-hadoop/dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/clean/oozie_app/workflow.xml

<workflow-app name="clean graph" xmlns="uri:oozie:workflow:0.5">

    <parameters>
        <property>
            <name>graphInputPath</name>
            <description>the input path to read graph content</description>
        </property>
        <property>
            <name>graphOutputPath</name>
            <description>the target path to store cleaned graph</description>
        </property>
        <property>
            <name>isLookupUrl</name>
            <description>the address of the lookUp service</description>
        </property>

        <property>
            <name>sparkDriverMemory</name>
            <description>memory for driver process</description>
        </property>
        <property>
            <name>sparkExecutorMemory</name>
            <description>memory for individual executor</description>
        </property>
        <property>
            <name>sparkExecutorCores</name>
            <description>number of cores used by single executor</description>
        </property>
        <property>
            <name>oozieActionShareLibForSpark2</name>
            <description>oozie action sharelib for spark 2.*</description>
        </property>
        <property>
            <name>spark2ExtraListeners</name>
            <value>com.cloudera.spark.lineage.NavigatorAppListener</value>
            <description>spark 2.* extra listeners classname</description>
        </property>
        <property>
            <name>spark2SqlQueryExecutionListeners</name>
            <value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
            <description>spark 2.* sql query execution listeners classname</description>
        </property>
        <property>
            <name>spark2YarnHistoryServerAddress</name>
            <description>spark 2.* yarn history server address</description>
        </property>
        <property>
            <name>spark2EventLogDir</name>
            <description>spark 2.* event log dir location</description>
        </property>
    </parameters>

	<start to="fork_clean_graph"/>

    <kill name="Kill">
        <message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
    </kill>

    <fork name="fork_clean_graph">
        <path start="clean_publication"/>
        <path start="clean_dataset"/>
        <path start="clean_otherresearchproduct"/>
        <path start="clean_software"/>
        <path start="clean_datasource"/>
        <path start="clean_organization"/>
        <path start="clean_project"/>
        <path start="clean_relation"/>
    </fork>

    <action name="clean_publication">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn</master>
            <mode>cluster</mode>
            <name>Clean publications</name>
            <class>eu.dnetlib.dhp.oa.graph.clean.CleanGraphSparkJob</class>
            <jar>dhp-graph-mapper-${projectVersion}.jar</jar>
            <spark-opts>
                --executor-cores=${sparkExecutorCores}
                --executor-memory=${sparkExecutorMemory}
                --driver-memory=${sparkDriverMemory}
                --conf spark.extraListeners=${spark2ExtraListeners}
                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
                --conf spark.sql.shuffle.partitions=7680
            </spark-opts>
            <arg>--inputPath</arg><arg>${graphInputPath}/publication</arg>
            <arg>--outputPath</arg><arg>${graphOutputPath}/publication</arg>
            <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Publication</arg>
            <arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
        </spark>
        <ok to="wait_clean"/>
        <error to="Kill"/>
    </action>

    <action name="clean_dataset">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn</master>
            <mode>cluster</mode>
            <name>Clean datasets</name>
            <class>eu.dnetlib.dhp.oa.graph.clean.CleanGraphSparkJob</class>
            <jar>dhp-graph-mapper-${projectVersion}.jar</jar>
            <spark-opts>
                --executor-cores=${sparkExecutorCores}
                --executor-memory=${sparkExecutorMemory}
                --driver-memory=${sparkDriverMemory}
                --conf spark.extraListeners=${spark2ExtraListeners}
                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
                --conf spark.sql.shuffle.partitions=7680
            </spark-opts>
            <arg>--inputPath</arg><arg>${graphInputPath}/dataset</arg>
            <arg>--outputPath</arg><arg>${graphOutputPath}/dataset</arg>
            <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Dataset</arg>
            <arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
        </spark>
        <ok to="wait_clean"/>
        <error to="Kill"/>
    </action>

    <action name="clean_otherresearchproduct">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn</master>
            <mode>cluster</mode>
            <name>Clean otherresearchproducts</name>
            <class>eu.dnetlib.dhp.oa.graph.clean.CleanGraphSparkJob</class>
            <jar>dhp-graph-mapper-${projectVersion}.jar</jar>
            <spark-opts>
                --executor-cores=${sparkExecutorCores}
                --executor-memory=${sparkExecutorMemory}
                --driver-memory=${sparkDriverMemory}
                --conf spark.extraListeners=${spark2ExtraListeners}
                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
                --conf spark.sql.shuffle.partitions=7680
            </spark-opts>
            <arg>--inputPath</arg><arg>${graphInputPath}/otherresearchproduct</arg>
            <arg>--outputPath</arg><arg>${graphOutputPath}/otherresearchproduct</arg>
            <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.OtherResearchProduct</arg>
            <arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
        </spark>
        <ok to="wait_clean"/>
        <error to="Kill"/>
    </action>

    <action name="clean_software">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn</master>
            <mode>cluster</mode>
            <name>Clean softwares</name>
            <class>eu.dnetlib.dhp.oa.graph.clean.CleanGraphSparkJob</class>
            <jar>dhp-graph-mapper-${projectVersion}.jar</jar>
            <spark-opts>
                --executor-cores=${sparkExecutorCores}
                --executor-memory=${sparkExecutorMemory}
                --driver-memory=${sparkDriverMemory}
                --conf spark.extraListeners=${spark2ExtraListeners}
                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
                --conf spark.sql.shuffle.partitions=7680
            </spark-opts>
            <arg>--inputPath</arg><arg>${graphInputPath}/software</arg>
            <arg>--outputPath</arg><arg>${graphOutputPath}/software</arg>
            <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Software</arg>
            <arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
        </spark>
        <ok to="wait_clean"/>
        <error to="Kill"/>
    </action>

    <action name="clean_datasource">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn</master>
            <mode>cluster</mode>
            <name>Clean datasources</name>
            <class>eu.dnetlib.dhp.oa.graph.clean.CleanGraphSparkJob</class>
            <jar>dhp-graph-mapper-${projectVersion}.jar</jar>
            <spark-opts>
                --executor-cores=${sparkExecutorCores}
                --executor-memory=${sparkExecutorMemory}
                --driver-memory=${sparkDriverMemory}
                --conf spark.extraListeners=${spark2ExtraListeners}
                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
                --conf spark.sql.shuffle.partitions=7680
            </spark-opts>
            <arg>--inputPath</arg><arg>${graphInputPath}/datasource</arg>
            <arg>--outputPath</arg><arg>${graphOutputPath}/datasource</arg>
            <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Datasource</arg>
            <arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
        </spark>
        <ok to="wait_clean"/>
        <error to="Kill"/>
    </action>

    <action name="clean_organization">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn</master>
            <mode>cluster</mode>
            <name>Clean organizations</name>
            <class>eu.dnetlib.dhp.oa.graph.clean.CleanGraphSparkJob</class>
            <jar>dhp-graph-mapper-${projectVersion}.jar</jar>
            <spark-opts>
                --executor-cores=${sparkExecutorCores}
                --executor-memory=${sparkExecutorMemory}
                --driver-memory=${sparkDriverMemory}
                --conf spark.extraListeners=${spark2ExtraListeners}
                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
                --conf spark.sql.shuffle.partitions=7680
            </spark-opts>
            <arg>--inputPath</arg><arg>${graphInputPath}/organization</arg>
            <arg>--outputPath</arg><arg>${graphOutputPath}/organization</arg>
            <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Organization</arg>
            <arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
        </spark>
        <ok to="wait_clean"/>
        <error to="Kill"/>
    </action>

    <action name="clean_project">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn</master>
            <mode>cluster</mode>
            <name>Clean projects</name>
            <class>eu.dnetlib.dhp.oa.graph.clean.CleanGraphSparkJob</class>
            <jar>dhp-graph-mapper-${projectVersion}.jar</jar>
            <spark-opts>
                --executor-cores=${sparkExecutorCores}
                --executor-memory=${sparkExecutorMemory}
                --driver-memory=${sparkDriverMemory}
                --conf spark.extraListeners=${spark2ExtraListeners}
                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
                --conf spark.sql.shuffle.partitions=7680
            </spark-opts>
            <arg>--inputPath</arg><arg>${graphInputPath}/project</arg>
            <arg>--outputPath</arg><arg>${graphOutputPath}/project</arg>
            <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Project</arg>
            <arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
        </spark>
        <ok to="wait_clean"/>
        <error to="Kill"/>
    </action>

    <action name="clean_relation">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn</master>
            <mode>cluster</mode>
            <name>Clean relations</name>
            <class>eu.dnetlib.dhp.oa.graph.clean.CleanGraphSparkJob</class>
            <jar>dhp-graph-mapper-${projectVersion}.jar</jar>
            <spark-opts>
                --executor-cores=${sparkExecutorCores}
                --executor-memory=${sparkExecutorMemory}
                --driver-memory=${sparkDriverMemory}
                --conf spark.extraListeners=${spark2ExtraListeners}
                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
                --conf spark.sql.shuffle.partitions=7680
            </spark-opts>
            <arg>--inputPath</arg><arg>${graphInputPath}/relation</arg>
            <arg>--outputPath</arg><arg>${graphOutputPath}/relation</arg>
            <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Relation</arg>
            <arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
        </spark>
        <ok to="wait_clean"/>
        <error to="Kill"/>
    </action>

    <join name="wait_clean" to="End"/>

    <end name="End"/>
</workflow-app>