2019-12-06 13:38:00 +01:00
|
|
|
<workflow-app name="Dedup Entities" xmlns="uri:oozie:workflow:0.5">
|
|
|
|
<parameters>
|
|
|
|
<property>
|
|
|
|
<name>sourcePath</name>
|
|
|
|
<description>the source path</description>
|
|
|
|
</property>
|
|
|
|
<property>
|
|
|
|
<name>entity</name>
|
|
|
|
<description>the entity that should be processed</description>
|
|
|
|
</property>
|
|
|
|
<property>
|
|
|
|
<name>dedupConf</name>
|
|
|
|
<description>the dedup Configuration</description>
|
|
|
|
</property>
|
|
|
|
<property>
|
|
|
|
<name>targetPath</name>
|
|
|
|
<description>the target 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>
|
|
|
|
<description>number of cores used by single executor</description>
|
|
|
|
</property>
|
|
|
|
</parameters>
|
|
|
|
|
2019-12-18 16:28:22 +01:00
|
|
|
<start to="CreateSimRels"/>
|
2019-12-06 13:38:00 +01:00
|
|
|
|
|
|
|
|
|
|
|
<kill name="Kill">
|
|
|
|
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
|
|
|
|
</kill>
|
|
|
|
|
|
|
|
<!-- <action name="DeleteTargetPath">-->
|
|
|
|
<!-- <fs>-->
|
|
|
|
<!-- <delete path='${targetPath}/${entity}_simrel'/>-->
|
|
|
|
<!-- <delete path='${targetPath}/${entity}_mergeRels'/>-->
|
|
|
|
<!-- </fs>-->
|
|
|
|
<!-- <ok to="CreateSimRels"/>-->
|
|
|
|
<!-- <error to="Kill"/>-->
|
|
|
|
<!-- </action>-->
|
|
|
|
|
|
|
|
<action name="CreateSimRels">
|
|
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
|
|
<job-tracker>${jobTracker}</job-tracker>
|
|
|
|
<name-node>${nameNode}</name-node>
|
|
|
|
<master>yarn-cluster</master>
|
|
|
|
<mode>cluster</mode>
|
|
|
|
<name>Create Similarity Relations</name>
|
|
|
|
<class>eu.dnetlib.dedup.SparkCreateSimRels</class>
|
|
|
|
<jar>dhp-dedup-${projectVersion}.jar</jar>
|
|
|
|
<spark-opts>--executor-memory ${sparkExecutorMemory} --executor-cores ${sparkExecutorCores}
|
|
|
|
--driver-memory=${sparkDriverMemory} --conf
|
|
|
|
spark.extraListeners="com.cloudera.spark.lineage.NavigatorAppListener" --conf
|
|
|
|
spark.sql.queryExecutionListeners="com.cloudera.spark.lineage.NavigatorQueryListener" --conf
|
|
|
|
spark.sql.warehouse.dir="/user/hive/warehouse"
|
|
|
|
</spark-opts>
|
|
|
|
<arg>-mt</arg><arg>yarn-cluster</arg>
|
|
|
|
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
|
|
|
|
<arg>--targetPath</arg><arg>${targetPath}</arg>
|
|
|
|
<arg>--entity</arg><arg>${entity}</arg>
|
|
|
|
<arg>--dedupConf</arg><arg>${dedupConf}</arg>
|
|
|
|
</spark>
|
|
|
|
<ok to="CreateConnectedComponents"/>
|
|
|
|
<error to="Kill"/>
|
|
|
|
</action>
|
|
|
|
|
|
|
|
|
|
|
|
<action name="CreateConnectedComponents">
|
|
|
|
<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>Create Connected Components</name>
|
|
|
|
<class>eu.dnetlib.dedup.SparkCreateConnectedComponent</class>
|
|
|
|
<jar>dhp-dedup-${projectVersion}.jar</jar>
|
|
|
|
<spark-opts>--executor-memory ${sparkExecutorMemory} --executor-cores ${sparkExecutorCores}
|
|
|
|
--driver-memory=${sparkDriverMemory} --conf
|
|
|
|
spark.extraListeners="com.cloudera.spark.lineage.NavigatorAppListener" --conf
|
|
|
|
spark.sql.queryExecutionListeners="com.cloudera.spark.lineage.NavigatorQueryListener" --conf
|
|
|
|
spark.sql.warehouse.dir="/user/hive/warehouse"
|
|
|
|
</spark-opts>
|
|
|
|
<arg>-mt</arg><arg>yarn-cluster</arg>
|
|
|
|
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
|
|
|
|
<arg>--targetPath</arg><arg>${targetPath}</arg>
|
|
|
|
<arg>--entity</arg><arg>${entity}</arg>
|
|
|
|
<arg>--dedupConf</arg><arg>${dedupConf}</arg>
|
|
|
|
</spark>
|
2019-12-11 15:43:24 +01:00
|
|
|
<ok to="CreateDedupRecord"/>
|
|
|
|
<error to="Kill"/>
|
|
|
|
</action>
|
|
|
|
|
|
|
|
<action name="CreateDedupRecord">
|
|
|
|
<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>
|
2019-12-18 16:28:22 +01:00
|
|
|
<name>Create Dedup Record</name>
|
2019-12-11 15:43:24 +01:00
|
|
|
<class>eu.dnetlib.dedup.SparkCreateDedupRecord</class>
|
|
|
|
<jar>dhp-dedup-${projectVersion}.jar</jar>
|
|
|
|
<spark-opts>--executor-memory ${sparkExecutorMemory} --executor-cores ${sparkExecutorCores}
|
|
|
|
--driver-memory=${sparkDriverMemory} --conf
|
|
|
|
spark.extraListeners="com.cloudera.spark.lineage.NavigatorAppListener" --conf
|
|
|
|
spark.sql.queryExecutionListeners="com.cloudera.spark.lineage.NavigatorQueryListener" --conf
|
|
|
|
spark.sql.warehouse.dir="/user/hive/warehouse"
|
|
|
|
</spark-opts>
|
|
|
|
<arg>-mt</arg><arg>yarn-cluster</arg>
|
|
|
|
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
|
|
|
|
<arg>--dedupPath</arg><arg>${dedupPath}</arg>
|
|
|
|
<arg>--entity</arg><arg>${entity}</arg>
|
|
|
|
<arg>--dedupConf</arg><arg>${dedupConf}</arg>
|
|
|
|
</spark>
|
2019-12-06 13:38:00 +01:00
|
|
|
<ok to="End"/>
|
|
|
|
<error to="Kill"/>
|
|
|
|
</action>
|
|
|
|
|
|
|
|
<end name="End"/>
|
|
|
|
</workflow-app>
|