alessia.bardi
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
dnet-hadoop/dhp-workflows/dhp-graph-provision/src/main/resources/eu/dnetlib/dhp/oa/provision/oozie_app/workflow.xml

652 lines
33 KiB
XML
Raw Blame History

<workflow-app name="index_infospace_graph" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>inputGraphRootPath</name>
<description>root location of input materialized graph</description>
</property>
<property>
<name>isLookupUrl</name>
<description>URL for the isLookup service</description>
</property>
<property>
<name>relPartitions</name>
<description>number or partitions for the relations Dataset</description>
</property>
<property>
<name>relationFilter</name>
<description>filter applied reading relations (by relClass)</description>
</property>
<property>
<name>sourceMaxRelations</name>
<description>maximum number of relations allowed for a each entity grouping by source</description>
</property>
<property>
<name>targetMaxRelations</name>
<description>maximum number of relations allowed for a each entity grouping by target</description>
</property>
<property>
<name>otherDsTypeId</name>
<description>mapping used to populate datasourceTypeUi field</description>
</property>
<property>
<name>format</name>
<description>metadata format name (DMF|TMF)</description>
</property>
<property>
<name>batchSize</name>
<description>number of records to be included in each indexing request</description>
</property>
<property>
<name>solrDeletionQuery</name>
<value>*:*</value>
<description>query used in the deleted by query operation</description>
</property>
<property>
<name>sparkDriverMemoryForJoining</name>
<description>memory for driver process</description>
</property>
<property>
<name>sparkExecutorMemoryForJoining</name>
<description>memory for individual executor</description>
</property>
<property>
<name>sparkExecutorCoresForJoining</name>
<description>number of cores used by single executor</description>
</property>
<property>
<name>sparkDriverMemoryForIndexing</name>
<description>memory for driver process</description>
</property>
<property>
<name>sparkExecutorMemoryForIndexing</name>
<description>memory for individual executor</description>
</property>
<property>
<name>sparkExecutorCoresForIndexing</name>
<description>number of cores used by single executor</description>
</property>
<property>
<name>oozieActionShareLibForSpark2</name>
<description>oozie action sharelib for spark 2.*</description>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
<description>spark 2.* extra listeners classname</description>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
<description>spark 2.* sql query execution listeners classname</description>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<description>spark 2.* yarn history server address</description>
</property>
<property>
<name>spark2EventLogDir</name>
<description>spark 2.* event log dir location</description>
</property>
<property>
<name>sparkNetworkTimeout</name>
<description>configures spark.network.timeout</description>
</property>
</parameters>
<global>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<configuration>
<property>
<name>oozie.action.sharelib.for.spark</name>
<value>${oozieActionShareLibForSpark2}</value>
</property>
</configuration>
</global>
<start to="resume_from"/>
<decision name="resume_from">
<switch>
<case to="prepare_relations">${wf:conf('resumeFrom') eq 'prepare_relations'}</case>
<case to="fork_join_related_entities">${wf:conf('resumeFrom') eq 'fork_join_related_entities'}</case>
<case to="fork_join_all_entities">${wf:conf('resumeFrom') eq 'fork_join_all_entities'}</case>
<case to="convert_to_xml">${wf:conf('resumeFrom') eq 'convert_to_xml'}</case>
<case to="drop_solr_collection">${wf:conf('resumeFrom') eq 'drop_solr_collection'}</case>
<case to="to_solr_index">${wf:conf('resumeFrom') eq 'to_solr_index'}</case>
<default to="prepare_relations"/>
</switch>
</decision>
<kill name="Kill">
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
</kill>
<action name="prepare_relations">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>PrepareRelations</name>
<class>eu.dnetlib.dhp.oa.provision.PrepareRelationsJob</class>
<jar>dhp-graph-provision-${projectVersion}.jar</jar>
<spark-opts>
--executor-cores=${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}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=3840
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${inputGraphRootPath}/relation</arg>
<arg>--outputPath</arg><arg>${workingDir}/relation</arg>
<arg>--sourceMaxRelations</arg><arg>${sourceMaxRelations}</arg>
<arg>--targetMaxRelations</arg><arg>${targetMaxRelations}</arg>
<arg>--relationFilter</arg><arg>${relationFilter}</arg>
<arg>--relPartitions</arg><arg>5000</arg>
</spark>
<ok to="fork_join_related_entities"/>
<error to="Kill"/>
</action>
<fork name="fork_join_related_entities">
<path start="join_relation_publication"/>
<path start="join_relation_dataset"/>
<path start="join_relation_otherresearchproduct"/>
<path start="join_relation_software"/>
<path start="join_relation_datasource"/>
<path start="join_relation_organization"/>
<path start="join_relation_project"/>
</fork>
<action name="join_relation_publication">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Join[relation.target = publication.id]</name>
<class>eu.dnetlib.dhp.oa.provision.CreateRelatedEntitiesJob_phase1</class>
<jar>dhp-graph-provision-${projectVersion}.jar</jar>
<spark-opts>
--executor-cores=${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}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=15000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${workingDir}/relation</arg>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/publication</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Publication</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_partial/publication</arg>
</spark>
<ok to="wait_joins"/>
<error to="Kill"/>
</action>
<action name="join_relation_dataset">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Join[relation.target = dataset.id]</name>
<class>eu.dnetlib.dhp.oa.provision.CreateRelatedEntitiesJob_phase1</class>
<jar>dhp-graph-provision-${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}
--conf spark.sql.shuffle.partitions=15000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${workingDir}/relation</arg>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/dataset</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Dataset</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_partial/dataset</arg>
</spark>
<ok to="wait_joins"/>
<error to="Kill"/>
</action>
<action name="join_relation_otherresearchproduct">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Join[relation.target = otherresearchproduct.id]</name>
<class>eu.dnetlib.dhp.oa.provision.CreateRelatedEntitiesJob_phase1</class>
<jar>dhp-graph-provision-${projectVersion}.jar</jar>
<spark-opts>
--executor-cores=${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}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=10000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${workingDir}/relation</arg>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/otherresearchproduct</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.OtherResearchProduct</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_partial/otherresearchproduct</arg>
</spark>
<ok to="wait_joins"/>
<error to="Kill"/>
</action>
<action name="join_relation_software">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Join[relation.target = software.id]</name>
<class>eu.dnetlib.dhp.oa.provision.CreateRelatedEntitiesJob_phase1</class>
<jar>dhp-graph-provision-${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}
--conf spark.sql.shuffle.partitions=5000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${workingDir}/relation</arg>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/software</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Software</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_partial/software</arg>
</spark>
<ok to="wait_joins"/>
<error to="Kill"/>
</action>
<action name="join_relation_datasource">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Join[relation.target = datasource.id]</name>
<class>eu.dnetlib.dhp.oa.provision.CreateRelatedEntitiesJob_phase1</class>
<jar>dhp-graph-provision-${projectVersion}.jar</jar>
<spark-opts>
--executor-cores=${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}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=5000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${workingDir}/relation</arg>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/datasource</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Datasource</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_partial/datasource</arg>
</spark>
<ok to="wait_joins"/>
<error to="Kill"/>
</action>
<action name="join_relation_organization">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Join[relation.target = organization.id]</name>
<class>eu.dnetlib.dhp.oa.provision.CreateRelatedEntitiesJob_phase1</class>
<jar>dhp-graph-provision-${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}
--conf spark.sql.shuffle.partitions=5000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${workingDir}/relation</arg>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/organization</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Organization</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_partial/organization</arg>
</spark>
<ok to="wait_joins"/>
<error to="Kill"/>
</action>
<action name="join_relation_project">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Join[relation.target = project.id]</name>
<class>eu.dnetlib.dhp.oa.provision.CreateRelatedEntitiesJob_phase1</class>
<jar>dhp-graph-provision-${projectVersion}.jar</jar>
<spark-opts>
--executor-cores=${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}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=5000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${workingDir}/relation</arg>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/project</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Project</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_partial/project</arg>
</spark>
<ok to="wait_joins"/>
<error to="Kill"/>
</action>
<join name="wait_joins" to="fork_join_all_entities"/>
<fork name="fork_join_all_entities">
<path start="join_publication_relations"/>
<path start="join_dataset_relations"/>
<path start="join_otherresearchproduct_relations"/>
<path start="join_software_relations"/>
<path start="join_datasource_relations"/>
<path start="join_organization_relations"/>
<path start="join_project_relations"/>
</fork>
<action name="join_publication_relations">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Join[publication.id = relatedEntity.source]</name>
<class>eu.dnetlib.dhp.oa.provision.CreateRelatedEntitiesJob_phase2</class>
<jar>dhp-graph-provision-${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}
--conf spark.sql.shuffle.partitions=15000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/publication</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Publication</arg>
<arg>--inputRelatedEntitiesPath</arg><arg>${workingDir}/join_partial</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_entities/publication</arg>
<arg>--numPartitions</arg><arg>30000</arg>
</spark>
<ok to="wait_join_phase2"/>
<error to="Kill"/>
</action>
<action name="join_dataset_relations">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Join[dataset.id = relatedEntity.source]</name>
<class>eu.dnetlib.dhp.oa.provision.CreateRelatedEntitiesJob_phase2</class>
<jar>dhp-graph-provision-${projectVersion}.jar</jar>
<spark-opts>
--executor-cores=${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}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=10000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/dataset</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Dataset</arg>
<arg>--inputRelatedEntitiesPath</arg><arg>${workingDir}/join_partial</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_entities/dataset</arg>
<arg>--numPartitions</arg><arg>20000</arg>
</spark>
<ok to="wait_join_phase2"/>
<error to="Kill"/>
</action>
<action name="join_otherresearchproduct_relations">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Join[otherresearchproduct.id = relatedEntity.source]</name>
<class>eu.dnetlib.dhp.oa.provision.CreateRelatedEntitiesJob_phase2</class>
<jar>dhp-graph-provision-${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}
--conf spark.sql.shuffle.partitions=10000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/otherresearchproduct</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.OtherResearchProduct</arg>
<arg>--inputRelatedEntitiesPath</arg><arg>${workingDir}/join_partial</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_entities/otherresearchproduct</arg>
<arg>--numPartitions</arg><arg>10000</arg>
</spark>
<ok to="wait_join_phase2"/>
<error to="Kill"/>
</action>
<action name="join_software_relations">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Join[software.id = relatedEntity.source]</name>
<class>eu.dnetlib.dhp.oa.provision.CreateRelatedEntitiesJob_phase2</class>
<jar>dhp-graph-provision-${projectVersion}.jar</jar>
<spark-opts>
--executor-cores=${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}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=5000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/software</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Software</arg>
<arg>--inputRelatedEntitiesPath</arg><arg>${workingDir}/join_partial</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_entities/software</arg>
<arg>--numPartitions</arg><arg>10000</arg>
</spark>
<ok to="wait_join_phase2"/>
<error to="Kill"/>
</action>
<action name="join_datasource_relations">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Join[datasource.id = relatedEntity.source]</name>
<class>eu.dnetlib.dhp.oa.provision.CreateRelatedEntitiesJob_phase2</class>
<jar>dhp-graph-provision-${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}
--conf spark.sql.shuffle.partitions=8000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/datasource</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Datasource</arg>
<arg>--inputRelatedEntitiesPath</arg><arg>${workingDir}/join_partial</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_entities/datasource</arg>
<arg>--numPartitions</arg><arg>1000</arg>
</spark>
<ok to="wait_join_phase2"/>
<error to="Kill"/>
</action>
<action name="join_organization_relations">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Join[organization.id = relatedEntity.source]</name>
<class>eu.dnetlib.dhp.oa.provision.CreateRelatedEntitiesJob_phase2</class>
<jar>dhp-graph-provision-${projectVersion}.jar</jar>
<spark-opts>
--executor-cores=${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}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=10000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/organization</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Organization</arg>
<arg>--inputRelatedEntitiesPath</arg><arg>${workingDir}/join_partial</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_entities/organization</arg>
<arg>--numPartitions</arg><arg>20000</arg>
</spark>
<ok to="wait_join_phase2"/>
<error to="Kill"/>
</action>
<action name="join_project_relations">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Join[project.id = relatedEntity.source]</name>
<class>eu.dnetlib.dhp.oa.provision.CreateRelatedEntitiesJob_phase2</class>
<jar>dhp-graph-provision-${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}
--conf spark.sql.shuffle.partitions=5000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/project</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Project</arg>
<arg>--inputRelatedEntitiesPath</arg><arg>${workingDir}/join_partial</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_entities/project</arg>
<arg>--numPartitions</arg><arg>10000</arg>
</spark>
<ok to="wait_join_phase2"/>
<error to="Kill"/>
</action>
<join name="wait_join_phase2" to="convert_to_xml"/>
<action name="convert_to_xml">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>convert_to_xml</name>
<class>eu.dnetlib.dhp.oa.provision.XmlConverterJob</class>
<jar>dhp-graph-provision-${projectVersion}.jar</jar>
<spark-opts>
--executor-cores=${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}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=3840
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputPath</arg><arg>${workingDir}/join_entities</arg>
<arg>--outputPath</arg><arg>${workingDir}/xml</arg>
<arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
<arg>--otherDsTypeId</arg><arg>${otherDsTypeId}</arg>
</spark>
<ok to="should_index"/>
<error to="Kill"/>
</action>
<decision name="should_index">
<switch>
<case to="drop_solr_collection">${wf:conf('shouldIndex') eq 'true'}</case>
<case to="End">${wf:conf('shouldIndex') eq 'false'}</case>
<default to="drop_solr_collection"/>
</switch>
</decision>
<action name="drop_solr_collection">
<java>
<main-class>eu.dnetlib.dhp.oa.provision.SolrAdminApplication</main-class>
<arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
<arg>--format</arg><arg>${format}</arg>
<arg>--action</arg><arg>DELETE_BY_QUERY</arg>
<arg>--query</arg><arg>${solrDeletionQuery}</arg>
<arg>--commit</arg><arg>true</arg>
</java>
<ok to="to_solr_index"/>
<error to="Kill"/>
</action>
<action name="to_solr_index">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>to_solr_index</name>
<class>eu.dnetlib.dhp.oa.provision.XmlIndexingJob</class>
<jar>dhp-graph-provision-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${sparkExecutorMemoryForIndexing}
--driver-memory=${sparkDriverMemoryForIndexing}
--conf spark.dynamicAllocation.enabled=true
--conf spark.dynamicAllocation.maxExecutors=${sparkExecutorCoresForIndexing}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.speculation=false
--conf spark.hadoop.mapreduce.map.speculative=false
--conf spark.hadoop.mapreduce.reduce.speculative=false
</spark-opts>
<arg>--inputPath</arg><arg>${workingDir}/xml</arg>
<arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
<arg>--format</arg><arg>${format}</arg>
<arg>--batchSize</arg><arg>${batchSize}</arg>
</spark>
<ok to="commit_solr_collection"/>
<error to="Kill"/>
</action>
<action name="commit_solr_collection">
<java>
<main-class>eu.dnetlib.dhp.oa.provision.SolrAdminApplication</main-class>
<arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
<arg>--format</arg><arg>${format}</arg>
<arg>--action</arg><arg>COMMIT</arg>
</java>
<ok to="End"/>
<error to="Kill"/>
</action>
<end name="End"/>
</workflow-app>
Reference in New Issue View Git Blame Copy Permalink