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

<workflow-app name="preprocess 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>
        <!-- Crossref Parameters -->
        <property>
            <name>inputPathCrossref</name>
            <description>the Crossref input path</description>
        </property>
        <property>
            <name>crossrefTimestamp</name>
            <description>Timestamp for the Crossref incremental Harvesting</description>
        </property>
        <property>
            <name>esServer</name>
            <description>elasticsearch server url for the Crossref Harvesting</description>
        </property>
        <property>
            <name>esIndex</name>
            <description>elasticsearch index name for the Crossref Harvesting</description>
        </property>
        <!--    MAG Parameters    -->
        <property>
            <name>MAGDumpPath</name>
            <description>the MAG dump working path</description>
        </property>

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

        <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="ImportCrossRef"/>

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

    <action name="ImportCrossRef">
        <java>
            <main-class>eu.dnetlib.doiboost.crossref.CrossrefImporter</main-class>
            <arg>--targetPath</arg><arg>${inputPathCrossref}/index_update</arg>
            <arg>--namenode</arg><arg>${nameNode}</arg>
            <arg>--esServer</arg><arg>${esServer}</arg>
            <arg>--esIndex</arg><arg>${esIndex}</arg>
            <arg>--timestamp</arg><arg>${crossrefTimestamp}</arg>
        </java>
        <ok to="GenerateCrossrefDataset"/>
        <error to="Kill"/>
    </action>


    <!-- CROSSREF SECTION -->

    <action name="GenerateCrossrefDataset">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn-cluster</master>
            <mode>cluster</mode>
            <name>GenerateCrossrefDataset</name>
            <class>eu.dnetlib.doiboost.crossref.CrossrefDataset</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}${spark2EventLogDir}
            </spark-opts>
            <arg>--workingPath</arg><arg>${inputPathCrossref}</arg>
            <arg>--master</arg><arg>yarn-cluster</arg>
        </spark>
        <ok to="RenameDataset"/>
        <error to="Kill"/>
    </action>

    <action name="RenameDataset">
        <fs>
            <delete path="${inputPathCrossref}/crossref_ds"/>
            <move source="${inputPathCrossref}/crossref_ds_updated"
                  target="${inputPathCrossref}/crossref_ds"/>
        </fs>
        <ok to="ResetMagWorkingPath"/>
        <error to="Kill"/>
    </action>



    <!-- MAG SECTION -->
    <action name="ResetMagWorkingPath">
        <fs>
            <delete path="${inputPathMAG}/dataset"/>
            <delete path="${inputPathMAG}/process"/>
        </fs>
        <ok to="ConvertMagToDataset"/>
        <error to="Kill"/>
    </action>

    <action name="ConvertMagToDataset">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <master>yarn-cluster</master>
            <mode>cluster</mode>
            <name>Convert Mag to Dataset</name>
            <class>eu.dnetlib.doiboost.mag.SparkImportMagIntoDataset</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}${spark2EventLogDir}
            </spark-opts>
            <arg>--sourcePath</arg><arg>${MAGDumpPath}</arg>
            <arg>--targetPath</arg><arg>${inputPathMAG}/dataset</arg>
            <arg>--master</arg><arg>yarn-cluster</arg>
        </spark>
        <ok to="PreprocessORCID"/>
        <error to="Kill"/>
    </action>



    <!--  ORCID  SECTION -->
    <action name="PreprocessORCID">
        <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.SparkPreprocessORCID</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}${spark2EventLogDir}
            </spark-opts>
            <arg>--sourcePath</arg><arg>${inputPathOrcid}</arg>
            <arg>--workingPath</arg><arg>${workingPathOrcid}</arg>
            <arg>--master</arg><arg>yarn-cluster</arg>
        </spark>
        <ok to="End"/>
        <error to="Kill"/>
    </action>


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