dnet-hadoop/dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/dump/wf/subworkflows/complete/oozie_app/workflow.xml

539 lines
25 KiB
XML

<workflow-app name="sub-dump_complete" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>sourcePath</name>
<description>the source path</description>
</property>
<property>
<name>outputPath</name>
<description>the output path</description>
</property>
<property>
<name>resultAggregation</name>
<description>true if all the result type have to be dumped under result. false otherwise</description>
</property>
<property>
<name>organizationCommunityMap</name>
<description>the organization community map</description>
</property>
<property>
<name>hiveDbName</name>
<description>the target hive database name</description>
</property>
<property>
<name>hiveJdbcUrl</name>
<description>hive server jdbc url</description>
</property>
<property>
<name>hiveMetastoreUris</name>
<description>hive server metastore URIs</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>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>
</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>${oozieActionShareLibForSpark2}</value>
</property>
</configuration>
</global>
<start to="fork_dump" />
<fork name="fork_dump">
<path start="dump_publication"/>
<path start="dump_dataset"/>
<path start="dump_orp"/>
<path start="dump_software"/>
<path start="dump_organization"/>
<path start="dump_project"/>
<path start="dump_datasource"/>
<path start="select_relation"/>
</fork>
<action name="dump_publication">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table publication </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkDumpEntitiesJob</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--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.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}/publication</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Publication</arg>
<arg>--outputPath</arg><arg>${workingDir}/result/publication</arg>
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
</spark>
<ok to="join_dump"/>
<error to="Kill"/>
</action>
<action name="dump_dataset">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table dataset </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkDumpEntitiesJob</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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}${spark2EventLogDir}
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}/dataset</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Dataset</arg>
<arg>--outputPath</arg><arg>${workingDir}/result/dataset</arg>
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
</spark>
<ok to="join_dump"/>
<error to="Kill"/>
</action>
<action name="dump_orp">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table ORP </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkDumpEntitiesJob</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--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.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}/otherresearchproduct</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.OtherResearchProduct</arg>
<arg>--outputPath</arg><arg>${workingDir}/result/otherresearchproduct</arg>
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
</spark>
<ok to="join_dump"/>
<error to="Kill"/>
</action>
<action name="dump_software">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table software </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkDumpEntitiesJob</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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}${spark2EventLogDir}
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}/software</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Software</arg>
<arg>--outputPath</arg><arg>${workingDir}/result/software</arg>
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
</spark>
<ok to="join_dump"/>
<error to="Kill"/>
</action>
<action name="dump_organization">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table organization </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkDumpEntitiesJob</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--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.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}/organization</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Organization</arg>
<arg>--outputPath</arg><arg>${outputPath}/organization</arg>
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
</spark>
<ok to="join_dump"/>
<error to="Kill"/>
</action>
<action name="dump_project">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table project </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkDumpEntitiesJob</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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}${spark2EventLogDir}
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}/project</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Project</arg>
<arg>--outputPath</arg><arg>${outputPath}/project</arg>
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
</spark>
<ok to="join_dump"/>
<error to="Kill"/>
</action>
<action name="dump_datasource">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table datasource </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkDumpEntitiesJob</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--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.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}/datasource</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Datasource</arg>
<arg>--outputPath</arg><arg>${outputPath}/datasource</arg>
<arg>--communityMapPath</arg><arg>${workingDir}/communityMap</arg>
</spark>
<ok to="join_dump"/>
<error to="Kill"/>
</action>
<action name="select_relation">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Select valid table relation </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkSelectValidRelationsJob</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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}${spark2EventLogDir}
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
--conf spark.sql.shuffle.partitions=3840
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
<arg>--outputPath</arg><arg>${workingDir}/validrelation</arg>
</spark>
<ok to="dump_relation"/>
<error to="Kill"/>
</action>
<action name="dump_relation">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table relation </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkDumpRelationJob</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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}${spark2EventLogDir}
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${workingDir}/validrelation</arg>
<arg>--outputPath</arg><arg>${workingDir}/relation/relation</arg>
<arg>--removeSet</arg><arg>${removeSet}</arg>
</spark>
<ok to="join_dump"/>
<error to="Kill"/>
</action>
<join name="join_dump" to="fork_context"/>
<fork name="fork_context">
<path start="create_entities_fromcontext"/>
<path start="create_relation_fromcontext"/>
<path start="create_relation_fromorgs"/>
</fork>
<action name="create_entities_fromcontext">
<java>
<main-class>eu.dnetlib.dhp.oa.graph.dump.complete.CreateContextEntities</main-class>
<arg>--hdfsPath</arg><arg>${outputPath}/communities_infrastructures/communities_infrastructure.json.gz</arg>
<arg>--nameNode</arg><arg>${nameNode}</arg>
<arg>--isLookUpUrl</arg><arg>${isLookUpUrl}</arg>
</java>
<ok to="join_context"/>
<error to="Kill"/>
</action>
<action name="create_relation_fromcontext">
<java>
<main-class>eu.dnetlib.dhp.oa.graph.dump.complete.CreateContextRelation</main-class>
<arg>--hdfsPath</arg><arg>${workingDir}/relation/context</arg>
<arg>--nameNode</arg><arg>${nameNode}</arg>
<arg>--isLookUpUrl</arg><arg>${isLookUpUrl}</arg>
</java>
<ok to="join_context"/>
<error to="Kill"/>
</action>
<action name="create_relation_fromorgs">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table relation </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkOrganizationRelation</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--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.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}/relation</arg>
<arg>--outputPath</arg><arg>${workingDir}/relation/contextOrg</arg>
<arg>--organizationCommunityMap</arg><arg>${organizationCommunityMap}</arg>
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
</spark>
<ok to="join_context"/>
<error to="Kill"/>
</action>
<join name="join_context" to="fork_extract_relations"/>
<fork name="fork_extract_relations">
<path start="rels_from_pubs"/>
<path start="rels_from_dats"/>
<path start="rels_from_orp"/>
<path start="rels_from_sw"/>
</fork>
<action name="rels_from_pubs">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Extract Relations from publication </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkExtractRelationFromEntities</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--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.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}/publication</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Publication</arg>
<arg>--outputPath</arg><arg>${workingDir}/relation/publication</arg>
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
</spark>
<ok to="join_extract_relations"/>
<error to="Kill"/>
</action>
<action name="rels_from_dats">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table dataset </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkExtractRelationFromEntities</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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}${spark2EventLogDir}
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}/dataset</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Dataset</arg>
<arg>--outputPath</arg><arg>${workingDir}/relation/dataset</arg>
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
</spark>
<ok to="join_extract_relations"/>
<error to="Kill"/>
</action>
<action name="rels_from_orp">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table ORP </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkExtractRelationFromEntities</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--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.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}/otherresearchproduct</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.OtherResearchProduct</arg>
<arg>--outputPath</arg><arg>${workingDir}/relation/orp</arg>
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
</spark>
<ok to="join_extract_relations"/>
<error to="Kill"/>
</action>
<action name="rels_from_sw">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table software </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkExtractRelationFromEntities</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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}${spark2EventLogDir}
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sourcePath}/software</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Software</arg>
<arg>--outputPath</arg><arg>${workingDir}/relation/software</arg>
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
</spark>
<ok to="join_extract_relations"/>
<error to="Kill"/>
</action>
<join name="join_extract_relations" to="collect_and_save"/>
<action name="collect_and_save">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Collect Results and Relations and put them in the right path </name>
<class>eu.dnetlib.dhp.oa.graph.dump.complete.SparkCollectAndSave</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--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.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${workingDir}</arg>
<arg>--outputPath</arg><arg>${outputPath}</arg>
<arg>--resultAggregation</arg><arg>${resultAggregation}</arg>
</spark>
<ok to="End"/>
<error to="Kill"/>
</action>
<kill name="Kill">
<message>Sub-workflow dump complete failed with error message ${wf:errorMessage()}
</message>
</kill>
<end name="End" />
</workflow-app>