forked from D-Net/dnet-hadoop
239 lines
12 KiB
XML
239 lines
12 KiB
XML
<workflow-app name="country_propagation" xmlns="uri:oozie:workflow:0.5">
|
|
<parameters>
|
|
<property>
|
|
<name>sourcePath</name>
|
|
<description>the source path</description>
|
|
</property>
|
|
<property>
|
|
<name>whitelist</name>
|
|
<description>the white list</description>
|
|
</property>
|
|
<property>
|
|
<name>allowedtypes</name>
|
|
<description>the allowed types</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>
|
|
<property>
|
|
<name>sparkExecutorNumber</name>
|
|
<description>number of executors used</description>
|
|
</property>
|
|
<property>
|
|
<name>writeUpdate</name>
|
|
<description>writes the information found for the update. No double check done if the information is already present</description>
|
|
</property>
|
|
<property>
|
|
<name>saveGraph</name>
|
|
<description>writes new version of the graph after the propagation step</description>
|
|
</property>
|
|
<property>
|
|
<name>outputPath</name>
|
|
<description>the path used to store temporary output files</description>
|
|
</property>
|
|
<property>
|
|
<name>preparedInfoPath</name>
|
|
<description>the path where prepared info have been stored</description>
|
|
</property>
|
|
|
|
</parameters>
|
|
|
|
<start to="prepare_datasource_country_association"/>
|
|
|
|
<kill name="Kill">
|
|
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
|
|
</kill>
|
|
|
|
<action name="prepare_datasource_country_association">
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
<master>yarn</master>
|
|
<mode>cluster</mode>
|
|
<name>PrepareDatasourceCountryAssociation</name>
|
|
<class>eu.dnetlib.dhp.countrypropagation.PrepareDatasourceCountryAssociation</class>
|
|
<jar>dhp-graph-propagation-${projectVersion}.jar</jar>
|
|
<spark-opts>
|
|
--executor-cores=${sparkExecutorCoresForJoining}
|
|
--executor-memory=${sparkExecutorMemoryForJoining}
|
|
--driver-memory=${sparkDriverMemoryForJoining}
|
|
--conf spark.extraListeners=${spark2ExtraListeners}
|
|
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
|
|
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
|
|
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
|
|
</spark-opts>
|
|
<arg>-mt</arg> <arg>yarn-cluster</arg>
|
|
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
|
|
<arg>--whitelist</arg><arg>${whitelist}</arg>
|
|
<arg>--allowedtypes</arg><arg>${allowedtypes}</arg>
|
|
<arg>--hive_metastore_uris</arg><arg>${hive_metastore_uris}</arg>
|
|
<arg>--outputPath</arg><arg>${workingDir}/country_propagation/preparedInfo</arg>
|
|
</spark>
|
|
<ok to="fork_join_apply_country_propagation"/>
|
|
<error to="Kill"/>
|
|
</action>
|
|
<fork name="fork_join_apply_country_propagation">
|
|
<!-- <path start="join_propagation_publication"/>-->
|
|
<!-- <path start="join_propagation_dataset"/>-->
|
|
<path start="join_propagation_otherresearchproduct"/>
|
|
<path start="join_propagation_software"/>
|
|
</fork>
|
|
|
|
<action name="join_propagation_publication">
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
<master>yarn</master>
|
|
<mode>cluster</mode>
|
|
<name>countryPropagationForPublications</name>
|
|
<class>eu.dnetlib.dhp.countrypropagation.SparkCountryPropagationJob2</class>
|
|
<jar>dhp-graph-propagation-${projectVersion}.jar</jar>
|
|
<spark-opts>
|
|
--executor-cores=${sparkExecutorCoresForJoining}
|
|
--executor-memory=${sparkExecutorMemoryForJoining}
|
|
--driver-memory=${sparkDriverMemoryForJoining}
|
|
--conf spark.extraListeners=${spark2ExtraListeners}
|
|
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
|
|
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
|
|
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
|
|
</spark-opts>
|
|
<arg>-mt</arg> <arg>yarn-cluster</arg>
|
|
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
|
|
<arg>--hive_metastore_uris</arg><arg>${hive_metastore_uris}</arg>
|
|
<arg>--writeUpdate</arg><arg>${writeUpdate}</arg>
|
|
<arg>--saveGraph</arg><arg>${saveGraph}</arg>
|
|
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Publication</arg>
|
|
<arg>--outputPath</arg><arg>${workingDir}/country_propagation/publication</arg>
|
|
<arg>--preparedInfoPath</arg><arg>${workingDir}/country_propagation/preparedInfo</arg>
|
|
</spark>
|
|
<ok to="end"/>
|
|
<error to="Kill"/>
|
|
</action>
|
|
|
|
<action name="join_propagation_dataset">
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
<master>yarn</master>
|
|
<mode>cluster</mode>
|
|
<name>countryPropagationForDataset</name>
|
|
<class>eu.dnetlib.dhp.countrypropagation.SparkCountryPropagationJob2</class>
|
|
<jar>dhp-graph-propagation-${projectVersion}.jar</jar>
|
|
<spark-opts>
|
|
--executor-cores=${sparkExecutorCoresForJoining}
|
|
--executor-memory=${sparkExecutorMemoryForJoining}
|
|
--driver-memory=${sparkDriverMemoryForJoining}
|
|
--conf spark.extraListeners=${spark2ExtraListeners}
|
|
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
|
|
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
|
|
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
|
|
</spark-opts>
|
|
<arg>-mt</arg> <arg>yarn-cluster</arg>
|
|
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
|
|
<arg>--hive_metastore_uris</arg><arg>${hive_metastore_uris}</arg>
|
|
<arg>--writeUpdate</arg><arg>${writeUpdate}</arg>
|
|
<arg>--saveGraph</arg><arg>${saveGraph}</arg>
|
|
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Dataset</arg>
|
|
<arg>--outputPath</arg><arg>${workingDir}/country_propagation/publication</arg>
|
|
<arg>--preparedInfoPath</arg><arg>${workingDir}/country_propagation/preparedInfo</arg>
|
|
</spark>
|
|
<ok to="end"/>
|
|
<error to="Kill"/>
|
|
</action>
|
|
|
|
<action name="join_propagation_otherresearchproduct">
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
<master>yarn</master>
|
|
<mode>cluster</mode>
|
|
<name>countryPropagationForORP</name>
|
|
<class>eu.dnetlib.dhp.countrypropagation.SparkCountryPropagationJob2</class>
|
|
<jar>dhp-graph-propagation-${projectVersion}.jar</jar>
|
|
<spark-opts>
|
|
--executor-cores=${sparkExecutorCoresForJoining}
|
|
--executor-memory=${sparkExecutorMemoryForJoining}
|
|
--driver-memory=${sparkDriverMemoryForJoining}
|
|
--conf spark.extraListeners=${spark2ExtraListeners}
|
|
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
|
|
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
|
|
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
|
|
</spark-opts>
|
|
<arg>-mt</arg> <arg>yarn-cluster</arg>
|
|
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
|
|
<arg>--hive_metastore_uris</arg><arg>${hive_metastore_uris}</arg>
|
|
<arg>--writeUpdate</arg><arg>${writeUpdate}</arg>
|
|
<arg>--saveGraph</arg><arg>${saveGraph}</arg>
|
|
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Otherresearchproduct</arg>
|
|
<arg>--outputPath</arg><arg>${workingDir}/country_propagation/publication</arg>
|
|
<arg>--preparedInfoPath</arg><arg>${workingDir}/country_propagation/preparedInfo</arg>
|
|
</spark>
|
|
<ok to="end"/>
|
|
<error to="Kill"/>
|
|
</action>
|
|
|
|
<action name="join_propagation_software">
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
<master>yarn</master>
|
|
<mode>cluster</mode>
|
|
<name>countryPropagationForSoftware</name>
|
|
<class>eu.dnetlib.dhp.countrypropagation.SparkCountryPropagationJob2</class>
|
|
<jar>dhp-graph-propagation-${projectVersion}.jar</jar>
|
|
<spark-opts>
|
|
--executor-cores=${sparkExecutorCoresForJoining}
|
|
--executor-memory=${sparkExecutorMemoryForJoining}
|
|
--driver-memory=${sparkDriverMemoryForJoining}
|
|
--conf spark.extraListeners=${spark2ExtraListeners}
|
|
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
|
|
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
|
|
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
|
|
</spark-opts>
|
|
<arg>-mt</arg> <arg>yarn-cluster</arg>
|
|
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
|
|
<arg>--hive_metastore_uris</arg><arg>${hive_metastore_uris}</arg>
|
|
<arg>--writeUpdate</arg><arg>${writeUpdate}</arg>
|
|
<arg>--saveGraph</arg><arg>${saveGraph}</arg>
|
|
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Software</arg>
|
|
<arg>--outputPath</arg><arg>${workingDir}/country_propagation/publication</arg>
|
|
<arg>--preparedInfoPath</arg><arg>${workingDir}/country_propagation/preparedInfo</arg>
|
|
</spark>
|
|
<ok to="end"/>
|
|
<error to="Kill"/>
|
|
</action>
|
|
|
|
<action name="CountryPropagation">
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
<job-tracker>${jobTracker}</job-tracker>
|
|
<name-node>${nameNode}</name-node>
|
|
<master>yarn-cluster</master>
|
|
<mode>cluster</mode>
|
|
<name>CountryPropagation</name>
|
|
<class>eu.dnetlib.dhp.countrypropagation.SparkCountryPropagationJob</class>
|
|
<jar>dhp-propagation-${projectVersion}.jar</jar>
|
|
<spark-opts>
|
|
--num-executors=${sparkExecutorNumber}
|
|
--executor-memory=${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}${spark2EventLogDir}
|
|
--conf spark.dynamicAllocation.enabled=true
|
|
--conf spark.dynamicAllocation.maxExecutors=${spark2MaxExecutors}
|
|
</spark-opts>
|
|
<arg>-mt</arg> <arg>yarn-cluster</arg>
|
|
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
|
|
<arg>--whitelist</arg><arg>${whitelist}</arg>
|
|
<arg>--allowedtypes</arg><arg>${allowedtypes}</arg>
|
|
<arg>--hive_metastore_uris</arg><arg>${hive_metastore_uris}</arg>
|
|
<arg>--writeUpdate</arg><arg>${writeUpdate}</arg>
|
|
<arg>--saveGraph</arg><arg>${saveGraph}</arg>
|
|
</spark>
|
|
<ok to="End"/>
|
|
<error to="Kill"/>
|
|
</action>
|
|
|
|
<end name="End"/>
|
|
</workflow-app> |