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

92 lines
3.2 KiB
XML

<workflow-app name="import MAG into HDFS" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>sourcePath</name>
<description>the working dir base path</description>
</property>
<property>
<name>targetPath</name>
<description>the working dir base path</description>
</property>
<property>
<name>workingPath</name>
<description>the working dir base 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>
<start to="ResetWorkingPath"/>
<kill name="Kill">
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
</kill>
<action name="ResetWorkingPath">
<fs>
<delete path='${workingPath}'/>
<mkdir path='${workingPath}'/>
</fs>
<ok to="ConvertMagToDataset"/>
<error to="Kill"/>
</action>
<action name="ConvertMagToDataset">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>Convert Mag to Dataset</name>
<class>eu.dnetlib.doiboost.mag.SparkImportMagIntoDataset</class>
<jar>dhp-doiboost-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
${sparkExtraOPT}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
<arg>--targetPath</arg><arg>${workingPath}</arg>
<arg>--master</arg><arg>yarn-cluster</arg>
</spark>
<ok to="PreprocessMag"/>
<error to="Kill"/>
</action>
<action name="PreprocessMag">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>Convert Mag to OAF Dataset</name>
<class>eu.dnetlib.doiboost.mag.SparkPreProcessMAG</class>
<jar>dhp-doiboost-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.sql.shuffle.partitions=3840
${sparkExtraOPT}
</spark-opts>
<arg>--sourcePath</arg><arg>${workingPath}</arg>
<arg>--workingPath</arg><arg>${workingPath}/process</arg>
<arg>--targetPath</arg><arg>${targetPath}</arg>
<arg>--master</arg><arg>yarn-cluster</arg>
</spark>
<ok to="End"/>
<error to="Kill"/>
</action>
<end name="End"/>
</workflow-app>