2020-01-24 09:58:55 +01:00
|
|
|
<workflow-app name="import_infospace_graph" xmlns="uri:oozie:workflow:0.5">
|
|
|
|
<parameters>
|
|
|
|
<property>
|
|
|
|
<name>sourcePath</name>
|
|
|
|
<description>the source path</description>
|
|
|
|
</property>
|
|
|
|
<property>
|
|
|
|
<name>targetPath</name>
|
|
|
|
<description>the source path</description>
|
|
|
|
</property>
|
|
|
|
<property>
|
|
|
|
<name>targetDir</name>
|
|
|
|
<description>the name of the 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>entities</name>
|
|
|
|
<description>the entities to be extracted</description>
|
|
|
|
</property>
|
|
|
|
</parameters>
|
|
|
|
|
2020-02-19 10:07:08 +01:00
|
|
|
<start to="DeleteTargetPath"/>
|
2020-01-24 09:58:55 +01:00
|
|
|
|
|
|
|
<kill name="Kill">
|
|
|
|
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
|
|
|
|
</kill>
|
2020-02-19 10:07:08 +01:00
|
|
|
<action name="DeleteTargetPath">
|
|
|
|
<fs>
|
|
|
|
<mkdir path="${targetPath}"/>
|
|
|
|
<mkdir path="${targetPath}/dataset"/>
|
|
|
|
<mkdir path="${targetPath}/publication"/>
|
|
|
|
<mkdir path="${targetPath}/unknown"/>
|
|
|
|
<mkdir path="${targetPath}/relation"/>
|
|
|
|
<delete path='${targetPath}/dataset/${targetDir}'/>
|
|
|
|
<delete path='${targetPath}/publication/${targetDir}'/>
|
|
|
|
<delete path='${targetPath}/unknown/${targetDir}'/>
|
|
|
|
<delete path='${targetPath}/relation/${targetDir}'/>
|
|
|
|
</fs>
|
|
|
|
<ok to="ExtractDLIEntities"/>
|
|
|
|
<error to="Kill"/>
|
|
|
|
</action>
|
2020-01-24 09:58:55 +01:00
|
|
|
|
2020-02-19 10:07:08 +01:00
|
|
|
<action name="ExtractDLIEntities">
|
2020-01-24 09:58:55 +01:00
|
|
|
<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>Extract ${entities}</name>
|
|
|
|
<class>eu.dnetlib.dhp.graph.scholexplorer.SparkExtractEntitiesJob</class>
|
|
|
|
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
|
|
|
|
<spark-opts>
|
|
|
|
--executor-memory ${sparkExecutorMemory}
|
2020-02-19 10:07:08 +01:00
|
|
|
--driver-memory=${sparkDriverMemory}
|
|
|
|
${sparkExtraOPT}
|
2020-01-24 09:58:55 +01:00
|
|
|
</spark-opts>
|
|
|
|
<arg>-mt</arg> <arg>yarn-cluster</arg>
|
|
|
|
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
|
|
|
|
<arg>--targetPath</arg><arg>${targetPath}</arg>
|
|
|
|
<arg>--targetDir</arg><arg>${targetDir}</arg>
|
|
|
|
<arg>--entities</arg><arg>${entities}</arg>
|
|
|
|
</spark>
|
|
|
|
<ok to="End"/>
|
|
|
|
<error to="Kill"/>
|
|
|
|
</action>
|
|
|
|
|
|
|
|
<end name="End"/>
|
|
|
|
</workflow-app>
|