dnet-hadoop/dhp-workflows/dhp-doiboost/src/main/resources/eu/dnetlib/dhp/doiboost/process/oozie_app/workflow.xml

<workflow-app name="Generate DOIBoost ActionSet" xmlns="uri:oozie:workflow:0.5">
    <parameters>
        <property>
            <name>sparkDriverMemory</name>
            <description>memory for driver process</description>
        </property>
        <property>
            <name>sparkExecutorMemory</name>
            <description>memory for individual executor</description>
        </property>
        <property>
            <name>sparkExecutorIntersectionMemory</name>
            <description>memory for individual executor</description>
        </property>

        <property>
            <name>sparkExecutorCores</name>
            <description>number of cores used by single executor</description>
        </property>
        <!-- Itersection Parameters -->
        <property>
            <name>workingPath</name>
            <description>the working Path</description>
        </property>

        <property>
            <name>hostedByMapPath</name>
            <description>the hostedByMap Path</description>
        </property>
        <property>
            <name>outputPath</name>
            <description>the Path of the sequence file action set</description>
        </property>


        <!-- Crossref Parameters -->
        <property>
            <name>inputPathCrossref</name>
            <description>the Crossref input path</description>
        </property>

        <!--    MAG Parameters    -->
         <property>
            <name>inputPathMAG</name>
            <description>the MAG working path</description>
        </property>


        <!--    UnpayWall Parameters    -->
        <property>
            <name>inputPathUnpayWall</name>
            <description>the UnpayWall working path</description>
        </property>

        <!--    ORCID Parameters    -->
        <property>
            <name>workingPathOrcid</name>
            <description>the ORCID working path</description>
        </property>

    </parameters>

    <global>
        <job-tracker>${jobTracker}</job-tracker>
        <name-node>${nameNode}</name-node>
        <configuration>
            <property>
                <name>oozie.action.sharelib.for.spark</name>
                <value>${oozieActionShareLibForSpark2}</value>
            </property>
        </configuration>
    </global>

    <start to="resume_from"/>

    <decision name="resume_from">
        <switch>
            <case to="ProcessMAG">${wf:conf('resumeFrom') eq 'PreprocessMag'}</case>
            <case to="ProcessUW">${wf:conf('resumeFrom') eq 'PreprocessUW'}</case>
            <case to="ProcessORCID">${wf:conf('resumeFrom') eq 'ProcessORCID'}</case>
            <case to="CreateDOIBoost">${wf:conf('resumeFrom') eq 'CreateDOIBoost'}</case>
            <case to="GenerateActionSet">${wf:conf('resumeFrom') eq 'GenerateActionSet'}</case>
            <default to="ConvertCrossrefToOAF"/>
        </switch>
    </decision>

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

    <action name="ConvertCrossrefToOAF">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn-cluster</master>
            <mode>cluster</mode>
            <name>ConvertCrossrefToOAF</name>
            <class>eu.dnetlib.doiboost.crossref.SparkMapDumpIntoOAF</class>
            <jar>dhp-doiboost-${projectVersion}.jar</jar>
            <spark-opts>
                --executor-memory=${sparkExecutorMemory}
                --executor-cores=${sparkExecutorCores}
                --driver-memory=${sparkDriverMemory}
                --conf spark.sql.shuffle.partitions=3840
                --conf spark.extraListeners=${spark2ExtraListeners}
                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
            </spark-opts>
            <arg>--sourcePath</arg><arg>${inputPathCrossref}/crossref_ds</arg>
            <arg>--targetPath</arg><arg>${workingPath}</arg>
            <arg>--master</arg><arg>yarn-cluster</arg>
        </spark>
        <ok to="ProcessMAG"/>
        <error to="Kill"/>
    </action>

    <action name="ProcessMAG">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn-cluster</master>
            <mode>cluster</mode>
            <name>Convert Mag to OAF Dataset</name>
            <class>eu.dnetlib.doiboost.mag.SparkProcessMAG</class>
            <jar>dhp-doiboost-${projectVersion}.jar</jar>
            <spark-opts>
                --executor-memory=${sparkExecutorIntersectionMemory}
                --executor-cores=${sparkExecutorCores}
                --driver-memory=${sparkDriverMemory}
                --conf spark.sql.shuffle.partitions=3840
                --conf spark.extraListeners=${spark2ExtraListeners}
                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
            </spark-opts>
            <arg>--sourcePath</arg><arg>${inputPathMAG}/dataset</arg>
            <arg>--workingPath</arg><arg>${inputPathMAG}/process_p</arg>
            <arg>--targetPath</arg><arg>${workingPath}</arg>
            <arg>--master</arg><arg>yarn-cluster</arg>
        </spark>
        <ok to="ProcessUW"/>
        <error to="Kill"/>
    </action>

    <!--  UnpayWall  SECTION -->

    <action name="ProcessUW">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn-cluster</master>
            <mode>cluster</mode>
            <name>Convert UnpayWall to Dataset</name>
            <class>eu.dnetlib.doiboost.uw.SparkMapUnpayWallToOAF</class>
            <jar>dhp-doiboost-${projectVersion}.jar</jar>
            <spark-opts>
                --executor-memory=${sparkExecutorMemory}
                --executor-cores=${sparkExecutorCores}
                --driver-memory=${sparkDriverMemory}
                --conf spark.sql.shuffle.partitions=3840
                --conf spark.sql.shuffle.partitions=3840
                --conf spark.extraListeners=${spark2ExtraListeners}
                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
            </spark-opts>
            <arg>--sourcePath</arg><arg>${inputPathUnpayWall}/uw_extracted</arg>
            <arg>--targetPath</arg><arg>${workingPath}/uwPublication</arg>
            <arg>--master</arg><arg>yarn-cluster</arg>
        </spark>
        <ok to="ProcessORCID"/>
        <error to="Kill"/>
    </action>

    <!--  ORCID  SECTION -->
    <action name="ProcessORCID">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn-cluster</master>
            <mode>cluster</mode>
            <name>Convert ORCID to Dataset</name>
            <class>eu.dnetlib.doiboost.orcid.SparkConvertORCIDToOAF</class>
            <jar>dhp-doiboost-${projectVersion}.jar</jar>
            <spark-opts>
                --executor-memory=${sparkExecutorMemory}
                --executor-cores=${sparkExecutorCores}
                --driver-memory=${sparkDriverMemory}
                --conf spark.sql.shuffle.partitions=3840
                --conf spark.extraListeners=${spark2ExtraListeners}
                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
            </spark-opts>
            <arg>--workingPath</arg><arg>${workingPathOrcid}</arg>
            <arg>--targetPath</arg><arg>${workingPath}/orcidPublication</arg>
            <arg>--master</arg><arg>yarn-cluster</arg>
        </spark>
        <ok to="CreateDOIBoost"/>
        <error to="Kill"/>
    </action>

    <!-- INTERSECTION SECTION-->
    <action name="CreateDOIBoost">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn-cluster</master>
            <mode>cluster</mode>
            <name>Create DOIBoost Infospace</name>
            <class>eu.dnetlib.doiboost.SparkGenerateDoiBoost</class>
            <jar>dhp-doiboost-${projectVersion}.jar</jar>
            <spark-opts>
                --executor-memory=${sparkExecutorIntersectionMemory}
                --executor-cores=${sparkExecutorCores}
                --driver-memory=${sparkDriverMemory}
                --conf spark.sql.shuffle.partitions=3840
                --conf spark.sql.shuffle.partitions=3840
                --conf spark.extraListeners=${spark2ExtraListeners}
                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
            </spark-opts>
            <arg>--hostedByMapPath</arg><arg>${hostedByMapPath}</arg>
            <arg>--affiliationPath</arg><arg>${inputPathMAG}/dataset/Affiliations</arg>
            <arg>--paperAffiliationPath</arg><arg>${inputPathMAG}/dataset/PaperAuthorAffiliations</arg>
            <arg>--workingPath</arg><arg>${workingPath}</arg>
            <arg>--master</arg><arg>yarn-cluster</arg>
        </spark>
        <ok to="GenerateActionSet"/>
        <error to="Kill"/>
    </action>


    <action name="GenerateActionSet">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn-cluster</master>
            <mode>cluster</mode>
            <name>Generate DOIBoost ActionSet</name>
            <class>eu.dnetlib.doiboost.SparkGenerateDOIBoostActionSet</class>
            <jar>dhp-doiboost-${projectVersion}.jar</jar>
            <spark-opts>
                --executor-memory=${sparkExecutorMemory}
                --executor-cores=${sparkExecutorCores}
                --driver-memory=${sparkDriverMemory}
                --conf spark.sql.shuffle.partitions=3840
                --conf spark.extraListeners=${spark2ExtraListeners}
                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
            </spark-opts>
            <arg>--dbPublicationPath</arg><arg>${workingPath}/doiBoostPublicationFiltered</arg>
            <arg>--dbDatasetPath</arg><arg>${workingPath}/crossrefDataset</arg>
            <arg>--crossRefRelation</arg><arg>${workingPath}/crossrefRelation</arg>
            <arg>--dbaffiliationRelationPath</arg><arg>${workingPath}/doiBoostPublicationAffiliation</arg>
            <arg>--dbOrganizationPath</arg><arg>${workingPath}/doiBoostOrganization</arg>
            <arg>--targetPath</arg><arg>${workingPath}/actionDataSet</arg>
            <arg>--sFilePath</arg><arg>${outputPath}</arg>
            <arg>--master</arg><arg>yarn-cluster</arg>
        </spark>
        <ok to="End"/>
        <error to="Kill"/>
    </action>

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