|
|
|
@ -1,3 +1,4 @@
|
|
|
|
|
|
|
|
|
|
<workflow-app name="UnresolvedEntities" xmlns="uri:oozie:workflow:0.5">
|
|
|
|
|
<parameters>
|
|
|
|
|
<property>
|
|
|
|
@ -49,6 +50,25 @@
|
|
|
|
|
</property>
|
|
|
|
|
</parameters>
|
|
|
|
|
|
|
|
|
|
<global>
|
|
|
|
|
<job-tracker>${jobTracker}</job-tracker>
|
|
|
|
|
<name-node>${nameNode}</name-node>
|
|
|
|
|
<configuration>
|
|
|
|
|
<property>
|
|
|
|
|
<name>mapreduce.job.queuename</name>
|
|
|
|
|
<value>${queueName}</value>
|
|
|
|
|
</property>
|
|
|
|
|
<property>
|
|
|
|
|
<name>oozie.launcher.mapred.job.queue.name</name>
|
|
|
|
|
<value>${oozieLauncherQueueName}</value>
|
|
|
|
|
</property>
|
|
|
|
|
<property>
|
|
|
|
|
<name>oozie.action.sharelib.for.spark</name>
|
|
|
|
|
<value>${oozieActionShareLibForSpark2}</value>
|
|
|
|
|
</property>
|
|
|
|
|
|
|
|
|
|
</configuration>
|
|
|
|
|
</global>
|
|
|
|
|
<start to="prepareInfo"/>
|
|
|
|
|
|
|
|
|
|
<kill name="Kill">
|
|
|
|
@ -79,7 +99,7 @@
|
|
|
|
|
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
|
|
|
|
|
</spark-opts>
|
|
|
|
|
<arg>--sourcePath</arg><arg>${bipScorePath}</arg>
|
|
|
|
|
<arg>--outputPath</arg><arg>${workingDir}/prepared/bip</arg>
|
|
|
|
|
<arg>--outputPath</arg><arg>${workingDir}/prepared</arg>
|
|
|
|
|
</spark>
|
|
|
|
|
<ok to="join"/>
|
|
|
|
|
<error to="Kill"/>
|
|
|
|
@ -102,7 +122,7 @@
|
|
|
|
|
<master>yarn</master>
|
|
|
|
|
<mode>cluster</mode>
|
|
|
|
|
<name>Produces the unresolved from FOS!</name>
|
|
|
|
|
<class>eu.dnetlib.dhp.actionmanager.bipfinder.SparkAtomicActionScoreJob</class>
|
|
|
|
|
<class>eu.dnetlib.dhp.actionmanager.createunresolvedentities.PrepareFOSSparkJob</class>
|
|
|
|
|
<jar>dhp-aggregation-${projectVersion}.jar</jar>
|
|
|
|
|
<spark-opts>
|
|
|
|
|
--executor-memory=${sparkExecutorMemory}
|
|
|
|
@ -115,8 +135,7 @@
|
|
|
|
|
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
|
|
|
|
|
</spark-opts>
|
|
|
|
|
<arg>--sourcePath</arg><arg>${workingDir}/input/fos</arg>
|
|
|
|
|
|
|
|
|
|
<arg>--outputPath</arg><arg>${workingDir}/prepared/fos</arg>
|
|
|
|
|
<arg>--outputPath</arg><arg>${workingDir}/prepared</arg>
|
|
|
|
|
|
|
|
|
|
</spark>
|
|
|
|
|
<ok to="join"/>
|
|
|
|
@ -132,7 +151,7 @@
|
|
|
|
|
<master>yarn</master>
|
|
|
|
|
<mode>cluster</mode>
|
|
|
|
|
<name>Saves the result produced for bip and fos by grouping results with the same id</name>
|
|
|
|
|
<class>eu.dnetlib.dhp.actionmanager.bipfinder.CollectAndSave</class>
|
|
|
|
|
<class>eu.dnetlib.dhp.actionmanager.createunresolvedentities.SparkSaveUnresolved</class>
|
|
|
|
|
<jar>dhp-aggregation-${projectVersion}.jar</jar>
|
|
|
|
|
<spark-opts>
|
|
|
|
|
--executor-memory=${sparkExecutorMemory}
|
|
|
|
@ -144,7 +163,7 @@
|
|
|
|
|
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
|
|
|
|
|
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
|
|
|
|
|
</spark-opts>
|
|
|
|
|
<arg>--inputPath</arg><arg>${workingDir}/prepared</arg>
|
|
|
|
|
<arg>--sourcePath</arg><arg>${workingDir}/prepared</arg>
|
|
|
|
|
<arg>--outputPath</arg><arg>${outputPath}</arg>
|
|
|
|
|
</spark>
|
|
|
|
|
<ok to="End"/>
|
|
|
|
|