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

110 lines
4.2 KiB
XML
Raw Normal View History

<workflow-app name="import MAG into HDFS" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>crossrefPublicationPath</name>
<description>the Crossref Publication Path</description>
</property>
<property>
<name>crossrefDatasetPath</name>
<description>the Crossref Dataset Path</description>
</property>
<property>
<name>uwPublicationPath</name>
<description>the UnpayWall Publication Path</description>
</property>
<property>
<name>magPublicationPath</name>
<description>the MAG Publication Path</description>
</property>
<property>
<name>orcidPublicationPath</name>
<description>the ORCID Publication Path</description>
</property>
<property>
<name>workingDirPath</name>
<description>the Working Path</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>
</parameters>
2020-05-22 20:51:42 +02:00
<start to="CreateDOIBoost"/>
<kill name="Kill">
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
</kill>
<action name="ResetWorkingPath">
<fs>
<delete path='${workingDirPath}'/>
<mkdir path='${workingDirPath}'/>
</fs>
<ok to="CreateDOIBoost"/>
<error to="Kill"/>
</action>
<action name="CreateDOIBoost">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>Create DOIBoost Infospace</name>
<class>eu.dnetlib.doiboost.SparkGenerateDoiBoost</class>
<jar>dhp-doiboost-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.sql.shuffle.partitions=3840
${sparkExtraOPT}
</spark-opts>
<arg>--crossrefPublicationPath</arg><arg>${crossrefPublicationPath}</arg>
<arg>--crossrefDatasetPath</arg><arg>${crossrefDatasetPath}</arg>
<arg>--uwPublicationPath</arg><arg>${uwPublicationPath}</arg>
<arg>--magPublicationPath</arg><arg>${magPublicationPath}</arg>
<arg>--orcidPublicationPath</arg><arg>${orcidPublicationPath}</arg>
<arg>--workingDirPath</arg><arg>${workingDirPath}</arg>
<arg>--master</arg><arg>yarn-cluster</arg>
</spark>
<ok to="End"/>
<error to="Kill"/>
</action>
<!-- <action name="PreprocessMag">-->
<!-- <spark xmlns="uri:oozie:spark-action:0.2">-->
<!-- <master>yarn-cluster</master>-->
<!-- <mode>cluster</mode>-->
<!-- <name>Convert Mag to Dataset</name>-->
<!-- <class>eu.dnetlib.doiboost.mag.SparkPreProcessMAG</class>-->
<!-- <jar>dhp-doiboost-${projectVersion}.jar</jar>-->
<!-- <spark-opts>-->
<!-- &#45;&#45;executor-memory=${sparkExecutorMemory}-->
<!-- &#45;&#45;executor-cores=${sparkExecutorCores}-->
<!-- &#45;&#45;driver-memory=${sparkDriverMemory}-->
<!-- &#45;&#45;conf spark.sql.shuffle.partitions=3840-->
<!-- ${sparkExtraOPT}-->
<!-- </spark-opts>-->
<!-- <arg>&#45;&#45;sourcePath</arg><arg>${sourcePath}</arg>-->
<!-- <arg>&#45;&#45;targetPath</arg><arg>${targetPath}</arg>-->
<!-- <arg>&#45;&#45;master</arg><arg>yarn-cluster</arg>-->
<!-- </spark>-->
<!-- <ok to="End"/>-->
<!-- <error to="Kill"/>-->
<!-- </action>-->
<end name="End"/>
</workflow-app>