2023-12-07 09:59:52 +01:00
|
|
|
<workflow-app name="community_to_result_propagation_project" xmlns="uri:oozie:workflow:0.5">
|
|
|
|
<parameters>
|
|
|
|
<property>
|
|
|
|
<name>sourcePath</name>
|
|
|
|
<description>the source path</description>
|
|
|
|
</property>
|
|
|
|
<property>
|
|
|
|
<name>baseURL</name>
|
|
|
|
<description>the base URL to use to select the right community APIs</description>
|
|
|
|
</property>
|
2024-01-17 16:50:00 +01:00
|
|
|
<property>
|
|
|
|
<name>startFrom</name>
|
|
|
|
<value>undelete</value>
|
|
|
|
</property>
|
2023-12-07 09:59:52 +01:00
|
|
|
</parameters>
|
|
|
|
|
|
|
|
<global>
|
|
|
|
<job-tracker>${jobTracker}</job-tracker>
|
|
|
|
<name-node>${nameNode}</name-node>
|
|
|
|
<configuration>
|
|
|
|
<property>
|
|
|
|
<name>oozie.action.sharelib.for.spark</name>
|
|
|
|
<value>${oozieActionShareLibForSpark2}</value>
|
|
|
|
</property>
|
|
|
|
</configuration>
|
|
|
|
</global>
|
|
|
|
|
2024-01-17 16:50:00 +01:00
|
|
|
<start to="startFrom"/>
|
|
|
|
|
|
|
|
<decision name="startFrom">
|
|
|
|
<switch>
|
|
|
|
<case to="prepare_result_communitylist">${wf:conf('startFrom') eq 'undelete'}</case>
|
|
|
|
|
|
|
|
<default to="reset_outputpath"/>
|
|
|
|
</switch>
|
|
|
|
</decision>
|
2023-12-07 09:59:52 +01:00
|
|
|
|
|
|
|
<kill name="Kill">
|
|
|
|
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
|
|
|
|
</kill>
|
2023-12-20 14:26:54 +01:00
|
|
|
<action name="reset_outputpath">
|
|
|
|
<fs>
|
|
|
|
<delete path="${workingDir}"/>
|
|
|
|
<mkdir path="${workingDir}"/>
|
|
|
|
</fs>
|
|
|
|
<ok to="prepare_result_communitylist"/>
|
|
|
|
<error to="Kill"/>
|
|
|
|
</action>
|
2023-12-07 09:59:52 +01:00
|
|
|
|
|
|
|
<action name="prepare_result_communitylist">
|
|
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
|
|
<master>yarn</master>
|
|
|
|
<mode>cluster</mode>
|
|
|
|
<name>Prepare-Community-Result-Organization</name>
|
|
|
|
<class>eu.dnetlib.dhp.resulttocommunityfromproject.PrepareResultCommunitySet</class>
|
|
|
|
<jar>dhp-enrichment-${projectVersion}.jar</jar>
|
|
|
|
<spark-opts>
|
|
|
|
--executor-cores=6
|
|
|
|
--executor-memory=5G
|
|
|
|
--conf spark.executor.memoryOverhead=3g
|
|
|
|
--conf spark.sql.shuffle.partitions=3284
|
|
|
|
--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.dynamicAllocation.maxExecutors=${spark2MaxExecutors}
|
|
|
|
</spark-opts>
|
|
|
|
<arg>--sourcePath</arg><arg>${sourcePath}/relation</arg>
|
|
|
|
<arg>--outputPath</arg><arg>${workingDir}/communitythroughproject/preparedInfo/resultCommunityList</arg>
|
|
|
|
<arg>--baseURL</arg><arg>${baseURL}</arg>
|
|
|
|
</spark>
|
|
|
|
<ok to="exec-propagation"/>
|
|
|
|
<error to="Kill"/>
|
|
|
|
</action>
|
|
|
|
|
|
|
|
<action name="exec-propagation">
|
|
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
|
|
<master>yarn</master>
|
|
|
|
<mode>cluster</mode>
|
|
|
|
<name>community2resultfromproject</name>
|
|
|
|
<class>eu.dnetlib.dhp.resulttocommunityfromproject.SparkResultToCommunityFromProject</class>
|
|
|
|
<jar>dhp-enrichment-${projectVersion}.jar</jar>
|
|
|
|
<spark-opts>
|
|
|
|
--executor-cores=6
|
|
|
|
--executor-memory=5G
|
|
|
|
--conf spark.executor.memoryOverhead=3g
|
|
|
|
--conf spark.sql.shuffle.partitions=3284
|
|
|
|
--driver-memory=${sparkDriverMemory}
|
|
|
|
--conf spark.extraListeners=${spark2ExtraListeners}
|
|
|
|
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
|
|
|
|
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
|
|
|
|
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
|
|
|
|
--conf spark.dynamicAllocation.maxExecutors=${spark2MaxExecutors}
|
|
|
|
</spark-opts>
|
|
|
|
<arg>--preparedInfoPath</arg><arg>${workingDir}/communitythroughproject/preparedInfo/resultCommunityList</arg>
|
|
|
|
<arg>--sourcePath</arg><arg>${sourcePath}/</arg>
|
|
|
|
<arg>--outputPath</arg><arg>${workingDir}/communitythroughproject/</arg>
|
|
|
|
</spark>
|
2024-01-15 17:49:54 +01:00
|
|
|
<ok to="move-results"/>
|
2023-12-07 09:59:52 +01:00
|
|
|
<error to="Kill"/>
|
|
|
|
</action>
|
|
|
|
|
2024-01-15 17:49:54 +01:00
|
|
|
<action name="move-results">
|
|
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
|
|
<master>yarn</master>
|
|
|
|
<mode>cluster</mode>
|
2024-01-17 16:50:00 +01:00
|
|
|
<name>move results</name>
|
2024-01-15 17:49:54 +01:00
|
|
|
<class>eu.dnetlib.dhp.MoveResult</class>
|
|
|
|
<jar>dhp-enrichment-${projectVersion}.jar</jar>
|
|
|
|
<spark-opts>
|
|
|
|
--executor-cores=6
|
|
|
|
--executor-memory=5G
|
|
|
|
--conf spark.executor.memoryOverhead=3g
|
|
|
|
--conf spark.sql.shuffle.partitions=3284
|
|
|
|
--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.dynamicAllocation.maxExecutors=${spark2MaxExecutors}
|
|
|
|
</spark-opts>
|
|
|
|
<arg>--sourcePath</arg><arg>${workingDir}/communitythroughproject/</arg>
|
|
|
|
<arg>--outputPath</arg><arg>${sourcePath}/</arg>
|
|
|
|
<!-- <arg>outputPath</arg><arg>/tmp/miriam/rescomm/</arg>-->
|
|
|
|
</spark>
|
|
|
|
<ok to="End"/>
|
|
|
|
<error to="Kill"/>
|
|
|
|
</action>
|
2023-12-07 09:59:52 +01:00
|
|
|
|
|
|
|
<end name="End"/>
|
|
|
|
|
|
|
|
</workflow-app>
|