dnet-hadoop/dhp-workflows/dhp-graph-provision-scholexplorer/src/main/resources/eu/dnetlib/dhp/sx/provision/oozie_app/workflow.xml

<workflow-app name="Materialize and Index graph to ElasticSearch" xmlns="uri:oozie:workflow:0.5">
    <parameters>
        <property>
            <name>workingDirPath</name>
            <description>the source path</description>
        </property>
        <property>
            <name>graphPath</name>
            <description>the graph path</description>
        </property>
        <property>
            <name>index</name>
            <description>the index name</description>
        </property>
        <property>
            <name>esCluster</name>
            <description>the Index cluster</description>
        </property>
        <property>
            <name>sparkDriverMemory</name>
            <description>memory for driver process</description>
        </property>
        <property>
            <name>sparkExecutorMemory</name>
            <description>memory for individual executor</description>
        </property>
    </parameters>

    <start to="DeleteTargetPath"/>

    <kill name="Kill">
        <message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
    </kill>
    <action name="DeleteTargetPath">
        <fs>
            <delete path='${workingDirPath}'/>
            <mkdir path='${workingDirPath}'/>
        </fs>
        <ok to="CalculateRelatedItem"/>
        <error to="Kill"/>
    </action>

    <action name="CalculateRelatedItem">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <job-tracker>${jobTracker}</job-tracker>
            <name-node>${nameNode}</name-node>
            <master>yarn-cluster</master>
            <mode>cluster</mode>
            <name>calculate for each ID the number of related Dataset, publication and Unknown</name>
            <class>eu.dnetlib.dhp.provision.SparkExtractRelationCount</class>
            <jar>dhp-graph-provision-scholexplorer-${projectVersion}.jar</jar>
            <spark-opts>--executor-memory ${sparkExecutorMemory} --executor-cores=${sparkExecutorCores} --driver-memory=${sparkDriverMemory} ${sparkExtraOPT}</spark-opts>
            <arg>-mt</arg> <arg>yarn-cluster</arg>
            <arg>--workingDirPath</arg><arg>${workingDirPath}</arg>
            <arg>--relationPath</arg><arg>${graphPath}/relation</arg>
        </spark>
        <ok to="generateSummary"/>
        <error to="Kill"/>
    </action>

    <action name="generateSummary">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <job-tracker>${jobTracker}</job-tracker>
            <name-node>${nameNode}</name-node>
            <master>yarn-cluster</master>
            <mode>cluster</mode>
            <name>generate Summary</name>
            <class>eu.dnetlib.dhp.provision.SparkGenerateSummaryIndex</class>
            <jar>dhp-graph-provision-scholexplorer-${projectVersion}.jar</jar>
            <spark-opts>--executor-memory ${sparkExecutorMemory} --executor-cores=${sparkExecutorCores} --driver-memory=${sparkDriverMemory}  --conf spark.sql.shuffle.partitions=4000 ${sparkExtraOPT}</spark-opts>
            <arg>-mt</arg> <arg>yarn-cluster</arg>
            <arg>--workingDirPath</arg><arg>${workingDirPath}</arg>
            <arg>--graphPath</arg><arg>${graphPath}</arg>
        </spark>
        <ok to="generateScholix"/>
        <error to="Kill"/>
    </action>

    <action name="generateScholix">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <job-tracker>${jobTracker}</job-tracker>
            <name-node>${nameNode}</name-node>
            <master>yarn-cluster</master>
            <mode>cluster</mode>
            <name>generate Scholix</name>
            <class>eu.dnetlib.dhp.provision.SparkGenerateScholixIndex</class>
            <jar>dhp-graph-provision-scholexplorer-${projectVersion}.jar</jar>
            <spark-opts>--executor-memory ${sparkExecutorMemory} --executor-cores=${sparkExecutorCores} --driver-memory=${sparkDriverMemory} --conf spark.sql.shuffle.partitions=4000 ${sparkExtraOPT}</spark-opts>
            <arg>-mt</arg> <arg>yarn-cluster</arg>
            <arg>--workingDirPath</arg><arg>${workingDirPath}</arg>
            <arg>--graphPath</arg><arg>${graphPath}</arg>
        </spark>
        <ok to="datasetToJson"/>
        <error to="Kill"/>
    </action>


    <action name="datasetToJson">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <job-tracker>${jobTracker}</job-tracker>
            <name-node>${nameNode}</name-node>
            <master>yarn-cluster</master>
            <mode>cluster</mode>
            <name>generate Scholix</name>
            <class>eu.dnetlib.dhp.provision.SparkConvertDatasetToJson</class>
            <jar>dhp-graph-provision-scholexplorer-${projectVersion}.jar</jar>
            <spark-opts>--executor-memory ${sparkExecutorMemory} --executor-cores=${sparkExecutorCores} --driver-memory=${sparkDriverMemory} --conf spark.sql.shuffle.partitions=4000 ${sparkExtraOPT}</spark-opts>
            <arg>-m</arg> <arg>yarn-cluster</arg>
            <arg>--workingPath</arg><arg>${workingDirPath}</arg>
        </spark>
        <ok to="End"/>
        <error to="Kill"/>
    </action>


    <action name="DropAndCreateIndex">
        <java>
            <job-tracker>${jobTracker}</job-tracker>
            <name-node>${nameNode}</name-node>
            <main-class>eu.dnetlib.dhp.provision.DropAndCreateESIndex</main-class>
            <arg>-i</arg><arg>${index}</arg>
            <arg>-c</arg><arg>${esCluster}</arg>
        </java>
        <ok to="indexSummary"/>
        <error to="Kill"/>
    </action>


    <action name="indexSummary">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <job-tracker>${jobTracker}</job-tracker>
            <name-node>${nameNode}</name-node>
            <master>yarn-cluster</master>
            <mode>cluster</mode>
            <name>index summary</name>
            <class>eu.dnetlib.dhp.provision.SparkIndexCollectionOnES</class>
            <jar>dhp-graph-provision-scholexplorer-${projectVersion}.jar</jar>
            <spark-opts>--executor-memory ${sparkExecutorMemory}  --driver-memory=${sparkDriverMemory} ${sparkExtraOPT} --conf spark.dynamicAllocation.maxExecutors="8" </spark-opts>
            <arg>-mt</arg> <arg>yarn-cluster</arg>
            <arg>--sourcePath</arg><arg>${workingDirPath}/summary_json</arg>
            <arg>--index</arg><arg>${index}_object</arg>
            <arg>--idPath</arg><arg>id</arg>
            <arg>--cluster</arg><arg>${esCluster}</arg>
        </spark>
        <ok to="indexScholix"/>
        <error to="Kill"/>
    </action>

    <action name="indexScholix">
        <spark xmlns="uri:oozie:spark-action:0.2">
            <job-tracker>${jobTracker}</job-tracker>
            <name-node>${nameNode}</name-node>
            <master>yarn-cluster</master>
            <mode>cluster</mode>
            <name>index scholix</name>
            <class>eu.dnetlib.dhp.provision.SparkIndexCollectionOnES</class>
            <jar>dhp-graph-provision-scholexplorer-${projectVersion}.jar</jar>
            <spark-opts>--executor-memory ${sparkExecutorMemory}  --driver-memory=${sparkDriverMemory} ${sparkExtraOPT} --conf spark.dynamicAllocation.maxExecutors="8" </spark-opts>
            <arg>-mt</arg> <arg>yarn-cluster</arg>
            <arg>--sourcePath</arg><arg>${workingDirPath}/scholix_json</arg>
            <arg>--index</arg><arg>${index}_scholix</arg>
            <arg>--idPath</arg><arg>identifier</arg>
            <arg>--cluster</arg><arg>${esCluster}</arg>
        </spark>
        <ok to="End"/>
        <error to="Kill"/>
    </action>

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