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

118 lines
4.8 KiB
XML

<workflow-app name="read Crossref dump from HDFS" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>crossrefDumpPath</name>
<description>the working dir base path</description>
</property>
<property>
<name>inputPathCrossref</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>
<value>2</value>
<description>number of cores used by single executor</description>
</property>
</parameters>
<start to="GenerateCrossrefDataset"/>
<kill name="Kill">
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
</kill>
<action name="ImportCrossRef">
<java>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<main-class>eu.dnetlib.doiboost.crossref.ExtractCrossrefRecords</main-class>
<arg>--hdfsServerUri</arg><arg>${nameNode}</arg>
<arg>--crossrefFileNameTarGz</arg><arg>${crossrefDumpPath}/crossref.tar.gz</arg>
<arg>--workingPath</arg><arg>${crossrefDumpPath}</arg>
<arg>--outputPath</arg><arg>${workingDir}/files/</arg>
</java>
<ok to="UnpackCrossrefEntries"/>
<error to="Kill"/>
</action>
<action name="UnpackCrossrefEntries">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>SparkGenerateCrossrefDataset</name>
<class>eu.dnetlib.doiboost.crossref.UnpackCrtossrefEntries</class>
<jar>dhp-doiboost-${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>--master</arg><arg>yarn-cluster</arg>
<arg>--sourcePath</arg><arg>${crossrefDumpPath}/files</arg>
<arg>--targetPath</arg><arg>${inputPathCrossref}/crossref_ds</arg>
</spark>
<ok to="GenerateCrossrefDataset"/>
<error to="Kill"/>
</action>
<action name="GenerateCrossrefDataset">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>SparkGenerateCrossrefDataset</name>
<class>eu.dnetlib.doiboost.crossref.GenerateCrossrefDataset</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}${spark2EventLogDir}
</spark-opts>
<arg>--master</arg><arg>yarn-cluster</arg>
<arg>--sourcePath</arg><arg>${inputPathCrossref}/crossref_ds</arg>
<arg>--targetPath</arg><arg>${inputPathCrossref}/crossref_ds_updates</arg>
</spark>
<ok to="End"/>
<error to="Kill"/>
</action>
<!-- <action name="removeFiles">-->
<!-- <fs>-->
<!-- <delete path="${workingDir}/files"/>-->
<!-- </fs>-->
<!-- <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="End"/>-->
<!-- <error to="Kill"/>-->
<!-- </action>-->
<end name="End"/>
</workflow-app>