changed the workflow to add the step for selecting the valid relations

This commit is contained in:
Miriam Baglioni 2021-04-01 16:52:39 +02:00
parent f93356f690
commit daabc370c5
1 changed files with 26 additions and 2 deletions

View File

@ -148,7 +148,7 @@
<path start="dump_organization"/>
<path start="dump_project"/>
<path start="dump_datasource"/>
<path start="dump_relation"/>
<path start="select_relation"/>
</fork>
<action name="dump_publication">
@ -333,6 +333,30 @@
<error to="Kill"/>
</action>
<action name="select_relation">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Select valid table relation </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkSelectValidRelationsJob</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
<arg>--outputPath</arg><arg>${workingDir}/validrelation</arg>
</spark>
<ok to="dump_relation"/>
<error to="Kill"/>
</action>
<action name="dump_relation">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
@ -350,7 +374,7 @@
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}/relation</arg>
<arg>--sourcePath</arg><arg>${workingDir}/validrelation</arg>
<arg>--outputPath</arg><arg>${workingDir}/relation/relation</arg>
</spark>
<ok to="join_dump"/>