BrBETA_dnet-hadoop/dhp-workflows/dhp-propagation/src/main/resources/eu/dnetlib/dhp/projecttoresult/oozie_app/workflow.xml

<workflow-app name="project_to_result_propagation" xmlns="uri:oozie:workflow:0.5">
    <parameters>
        <property>
            <name>sourcePath</name>
            <description>the source path</description>
        </property>
        <property>
            <name>allowedsemrels</name>
            <description>the allowed semantics </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>writeUpdate</name>
            <description>writes the information found for the update. No double check done if the information is already present</description>
        </property>
        <property>
            <name>saveGraph</name>
            <description>writes new version of the graph after the propagation step</description>
        </property>
    </parameters>

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

    <action name="reset-outputpath">
        <fs>
            <delete path='${workingDir}/projecttoresult_propagation/relation'/>
        </fs>
        <ok to="copy_relation"/>
        <error to="Kill"/>
    </action>
    <action name="copy_relation">
        <distcp xmlns="uri:oozie:distcp-action:0.2">
            <job-tracker>${jobTracker}</job-tracker>
            <name-node>${nameNode}</name-node>
            <arg>${nameNode}/${sourcePath}/relation</arg>
            <arg>${nameNode}/${workingDir}/projecttoresult_propagation/relation</arg>
        </distcp>
        <ok to="prepare_project_results_association"/>
        <error to="Kill"/>
    </action>

        <action name="prepare_project_results_association">
            <spark xmlns="uri:oozie:spark-action:0.2">
                <master>yarn</master>
                <mode>cluster</mode>
                <name>PrepareProjectResultsAssociation</name>
                <class>eu.dnetlib.dhp.projecttoresult.PrepareProjectResultsAssociation</class>
                <jar>dhp-propagation-${projectVersion}.jar</jar>
                <spark-opts>
                    --executor-cores=${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}${spark2EventLogDir}
                </spark-opts>
                <arg>--sourcePath</arg><arg>${sourcePath}/relation</arg>
                <arg>--allowedsemrels</arg><arg>${allowedsemrels}</arg>
                <arg>--hive_metastore_uris</arg><arg>${hive_metastore_uris}</arg>
                <arg>--potentialUpdatePath</arg><arg>${workingDir}/projecttoresult_propagation/preparedInfo/potentialUpdates</arg>
                <arg>--alreadyLinkedPath</arg><arg>${workingDir}/projecttoresult_propagation/preparedInfo/alreadyLinked</arg>
            </spark>
            <ok to="apply_propagation"/>
            <error to="Kill"/>
        </action>

        <action name="apply_propagation">
            <spark xmlns="uri:oozie:spark-action:0.2">
                <master>yarn</master>
                <mode>cluster</mode>
                <name>ProjectToResultPropagation</name>
                <class>eu.dnetlib.dhp.projecttoresult.SparkResultToProjectThroughSemRelJob3</class>
                <jar>dhp-propagation-${projectVersion}.jar</jar>
                <spark-opts>
                    --executor-cores=${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}${spark2EventLogDir}
                    --conf spark.dynamicAllocation.enabled=true
                    --conf spark.dynamicAllocation.maxExecutors=${spark2MaxExecutors}
                </spark-opts>
<!--                <arg>&#45;&#45;sourcePath</arg><arg>${sourcePath}/relation</arg>-->
                <arg>--writeUpdate</arg><arg>${writeUpdate}</arg>
                <arg>--saveGraph</arg><arg>${saveGraph}</arg>
                <arg>--hive_metastore_uris</arg><arg>${hive_metastore_uris}</arg>
                <arg>--outputPath</arg><arg>${workingDir}/projecttoresult_propagation/relation</arg>
                <arg>--potentialUpdatePath</arg><arg>${workingDir}/projecttoresult_propagation/preparedInfo/potentialUpdates</arg>
                <arg>--alreadyLinkedPath</arg><arg>${workingDir}/projecttoresult_propagation/preparedInfo/alreadyLinked</arg>
            </spark>
            <ok to="End"/>
            <error to="Kill"/>
        </action>

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