dhp-graph-dump/dump/src/main/resources/eu/dnetlib/dhp/oa/graph/dump/countryresults/oozie_app/workflow.xml

540 lines
24 KiB
XML

<workflow-app name="sub-dump_subset" 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>country</name>
<description>the country for which to produce the dump</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>${oozieActionShareLibForSpark2}</value>
</property>
</configuration>
</global>
<!-- <start to="save_community_map" />-->
<start to="save_community_map" />
<action name="save_community_map">
<java>
<main-class>eu.dnetlib.dhp.oa.graph.dump.SaveCommunityMap</main-class>
<arg>--outputPath</arg><arg>${workingDir}/communityMap</arg>
<arg>--nameNode</arg><arg>${nameNode}</arg>
<arg>--isLookUpUrl</arg><arg>${isLookUpUrl}</arg>
</java>
<ok to="find_results_for_country"/>
<error to="Kill"/>
</action>
<action name="find_results_for_country">
<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.country.SparkFindResultsRelatedToCountry</class>
<jar>dump-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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>${sourcePath}</arg>
<arg>--outputPath</arg><arg>${workingDir}/resultsInCountry</arg>
<arg>--country</arg><arg>${country}</arg>
</spark>
<ok to="fork_select"/>
<error to="Kill"/>
</action>
<fork name="fork_select">
<path start="select_publication"/>
<path start="select_dataset"/>
<path start="select_orp"/>
<path start="select_software"/>
</fork>
<action name="select_publication">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table publication </name>
<class>eu.dnetlib.dhp.oa.graph.dump.country.SparkFindResultWithCountry</class>
<jar>dump-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=7G
--executor-cores=2
--driver-memory=7G
--conf spark.sql.shuffle.partitions=3840
--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>${sourcePath}/publication</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Publication</arg>
<arg>--outputPath</arg><arg>${outputPath}</arg>
<arg>--resultWithCountry</arg><arg>${workingDir}/resultsInCountry</arg>
<arg>--resultType</arg><arg>publication</arg>
</spark>
<ok to="join_dump"/>
<error to="Kill"/>
</action>
<action name="select_dataset">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table dataset </name>
<class>eu.dnetlib.dhp.oa.graph.dump.country.SparkFindResultWithCountry</class>
<jar>dump-${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>${outputPath}</arg>
<arg>--resultType</arg><arg>dataset</arg>
<arg>--resultWithCountry</arg><arg>${workingDir}/resultsInCountry</arg>
</spark>
<ok to="join_dump"/>
<error to="Kill"/>
</action>
<action name="select_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.country.SparkFindResultWithCountry</class>
<jar>dump-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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>${sourcePath}/otherresearchproduct</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.OtherResearchProduct</arg>
<arg>--outputPath</arg><arg>${outputPath}</arg>
<arg>--resultType</arg><arg>otherresearchproduct</arg>
<arg>--resultWithCountry</arg><arg>${workingDir}/resultsInCountry</arg>
</spark>
<ok to="join_dump"/>
<error to="Kill"/>
</action>
<action name="select_software">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table software </name>
<class>eu.dnetlib.dhp.oa.graph.dump.country.SparkFindResultWithCountry</class>
<jar>dump-${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>${outputPath}</arg>
<arg>--resultType</arg><arg>software</arg>
<arg>--resultWithCountry</arg><arg>${workingDir}/resultsInCountry</arg>
</spark>
<ok to="join_dump"/>
<error to="Kill"/>
</action>
<join name="join_dump" to="fork_dump_community"/>
<action name="select_subset">
<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.subset.SparkSelectSubset</class>
<jar>dump-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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>${sourcePath}</arg>
<arg>--outputPath</arg><arg>${outputPath}</arg>
<arg>--removeSet</arg><arg>${removeSet}</arg>
</spark>
<ok to="fork_dump_community"/>
<error to="Kill"/>
</action>
<fork name="fork_dump_community">
<path start="dump_publication"/>
<path start="dump_dataset"/>
<path start="dump_orp"/>
<path start="dump_software"/>
</fork>
<action name="dump_publication">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table publication for community/funder related products</name>
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkDumpCommunityProducts</class>
<jar>dump-${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=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${outputPath}/original/publication</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Publication</arg>
<arg>--outputPath</arg><arg>${workingDir}/dump/publication</arg>
<arg>--communityMapPath</arg><arg>${workingDir}/communityMap</arg>
<arg>--dumpType</arg><arg>country</arg>
</spark>
<ok to="join_dump_comm"/>
<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 for community/funder related products</name>
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkDumpCommunityProducts</class>
<jar>dump-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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>${outputPath}/original/dataset</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Dataset</arg>
<arg>--outputPath</arg><arg>${workingDir}/dump/dataset</arg>
<arg>--communityMapPath</arg><arg>${workingDir}/communityMap</arg>
<arg>--dumpType</arg><arg>country</arg>
</spark>
<ok to="join_dump_comm"/>
<error to="Kill"/>
</action>
<action name="dump_orp">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Dump table ORP for community related products</name>
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkDumpCommunityProducts</class>
<jar>dump-${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=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${outputPath}/original/otherresearchproduct</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.OtherResearchProduct</arg>
<arg>--outputPath</arg><arg>${workingDir}/dump/otherresearchproduct</arg>
<arg>--communityMapPath</arg><arg>${workingDir}/communityMap</arg>
<arg>--dumpType</arg><arg>country</arg>
</spark>
<ok to="join_dump_comm"/>
<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 for community related products</name>
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkDumpCommunityProducts</class>
<jar>dump-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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>${outputPath}/original/software</arg>
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Software</arg>
<arg>--outputPath</arg><arg>${workingDir}/dump/software</arg>
<arg>--communityMapPath</arg><arg>${workingDir}/communityMap</arg>
<arg>--dumpType</arg><arg>country</arg>
</spark>
<ok to="join_dump_comm"/>
<error to="Kill"/>
</action>
<join name="join_dump_comm" to="prepareResultProject"/>
<action name="prepareResultProject">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Prepare association result subset of project info</name>
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkPrepareResultProject</class>
<jar>dump-${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}</arg>
<arg>--outputPath</arg><arg>${workingDir}/preparedInfo</arg>
</spark>
<ok to="fork_extendWithProject"/>
<error to="Kill"/>
</action>
<fork name="fork_extendWithProject">
<path start="extend_publication"/>
<path start="extend_dataset"/>
<path start="extend_orp"/>
<path start="extend_software"/>
</fork>
<action name="extend_publication">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Extend dumped publications with information about project</name>
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkUpdateProjectInfo</class>
<jar>dump-${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}/dump/publication</arg>
<arg>--outputPath</arg><arg>${outputPath}/dump/publication</arg>
<arg>--preparedInfoPath</arg><arg>${workingDir}/preparedInfo</arg>
</spark>
<ok to="join_extend"/>
<error to="Kill"/>
</action>
<action name="extend_dataset">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Extend dumped dataset with information about project</name>
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkUpdateProjectInfo</class>
<jar>dump-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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>${workingDir}/dump/dataset</arg>
<arg>--outputPath</arg><arg>${outputPath}/dump/dataset</arg>
<arg>--preparedInfoPath</arg><arg>${workingDir}/preparedInfo</arg>
</spark>
<ok to="join_extend"/>
<error to="Kill"/>
</action>
<action name="extend_orp">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Extend dumped ORP with information about project</name>
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkUpdateProjectInfo</class>
<jar>dump-${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}/dump/otherresearchproduct</arg>
<arg>--outputPath</arg><arg>${outputPath}/dump/otherresearchproduct</arg>
<arg>--preparedInfoPath</arg><arg>${workingDir}/preparedInfo</arg>
</spark>
<ok to="join_extend"/>
<error to="Kill"/>
</action>
<action name="extend_software">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Extend dumped software with information about project</name>
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkUpdateProjectInfo</class>
<jar>dump-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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>${workingDir}/dump/software</arg>
<arg>--outputPath</arg><arg>${outputPath}/dump/software</arg>
<arg>--preparedInfoPath</arg><arg>${workingDir}/preparedInfo</arg>
</spark>
<ok to="join_extend"/>
<error to="Kill"/>
</action>
<join name="join_extend" to="make_archive"/>
<action name="make_archive">
<java>
<main-class>eu.dnetlib.dhp.oa.graph.dump.MakeTar</main-class>
<arg>--hdfsPath</arg><arg>${outputPath}/tar</arg>
<arg>--nameNode</arg><arg>${nameNode}</arg>
<arg>--sourcePath</arg><arg>${outputPath}/dump</arg>
</java>
<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>