[Graph DUMP] removed not used sub workflow
This commit is contained in:
parent
f738acb85a
commit
c0dab69349
|
@ -81,7 +81,7 @@ public class SparkDumpFunderResults implements Serializable {
|
|||
|
||||
}).collect(Collectors.toList()).iterator(), Encoders.STRING())
|
||||
.distinct();
|
||||
tmp.foreach((ForeachFunction<String>) f -> log.info("Found Funder {}", f));
|
||||
|
||||
List<String> funderList = tmp.collectAsList();
|
||||
|
||||
funderList.forEach(funder -> {
|
||||
|
|
|
@ -1,30 +0,0 @@
|
|||
<configuration>
|
||||
<property>
|
||||
<name>jobTracker</name>
|
||||
<value>yarnRM</value>
|
||||
</property>
|
||||
<property>
|
||||
<name>nameNode</name>
|
||||
<value>hdfs://nameservice1</value>
|
||||
</property>
|
||||
<property>
|
||||
<name>oozie.use.system.libpath</name>
|
||||
<value>true</value>
|
||||
</property>
|
||||
<property>
|
||||
<name>hiveMetastoreUris</name>
|
||||
<value>thrift://iis-cdh5-test-m3.ocean.icm.edu.pl:9083</value>
|
||||
</property>
|
||||
<property>
|
||||
<name>hiveJdbcUrl</name>
|
||||
<value>jdbc:hive2://iis-cdh5-test-m3.ocean.icm.edu.pl:10000</value>
|
||||
</property>
|
||||
<property>
|
||||
<name>hiveDbName</name>
|
||||
<value>openaire</value>
|
||||
</property>
|
||||
<property>
|
||||
<name>oozie.launcher.mapreduce.user.classpath.first</name>
|
||||
<value>true</value>
|
||||
</property>
|
||||
</configuration>
|
|
@ -1,345 +0,0 @@
|
|||
<workflow-app name="sub_dump_community_funder_results" 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>communityMapPath</name>
|
||||
<description>the path to the community map</description>
|
||||
</property>
|
||||
<property>
|
||||
<name>selectedResults</name>
|
||||
<description>the path the the possible subset ot results to be dumped</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>${oozieActionShareLibForSpark2}</value>
|
||||
</property>
|
||||
|
||||
</configuration>
|
||||
</global>
|
||||
|
||||
<start to="fork_dump"/>
|
||||
|
||||
<kill name="Kill">
|
||||
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
|
||||
</kill>
|
||||
|
||||
|
||||
<fork name="fork_dump">
|
||||
<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>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>${selectedResults}/publication</arg>
|
||||
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Publication</arg>
|
||||
<arg>--outputPath</arg><arg>${workingDir}/dump/publication</arg>
|
||||
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
|
||||
<arg>--dumpType</arg><arg>${dumpType}</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 for community/funder related products</name>
|
||||
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkDumpCommunityProducts</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>${selectedResults}/dataset</arg>
|
||||
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Dataset</arg>
|
||||
<arg>--outputPath</arg><arg>${workingDir}/dump/dataset</arg>
|
||||
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
|
||||
</spark>
|
||||
<ok to="join_dump"/>
|
||||
<error to="Kill"/>
|
||||
</action>
|
||||
|
||||
<action name="dump_orp">
|
||||
<spark xmlns="uri:oozie:spark-action:0.2">
|
||||
<master>yarn</master>
|
||||
<mode>cluster</mode>
|
||||
<name>Dump table ORP for community related products</name>
|
||||
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkDumpCommunityProducts</class>
|
||||
<jar>dhp-graph-mapper-${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=${sparkSqlWarehouseDir}
|
||||
</spark-opts>
|
||||
<arg>--sourcePath</arg><arg>${selectedResults}/otherresearchproduct</arg>
|
||||
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.OtherResearchProduct</arg>
|
||||
<arg>--outputPath</arg><arg>${workingDir}/dump/otherresearchproduct</arg>
|
||||
<arg>--communityMapPath</arg><arg>${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 for community related products</name>
|
||||
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkDumpCommunityProducts</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>${selectedResults}/software</arg>
|
||||
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Software</arg>
|
||||
<arg>--outputPath</arg><arg>${workingDir}/dump/software</arg>
|
||||
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
|
||||
</spark>
|
||||
<ok to="join_dump"/>
|
||||
<error to="Kill"/>
|
||||
</action>
|
||||
|
||||
<join name="join_dump" to="prepareResultProject"/>
|
||||
|
||||
<action name="prepareResultProject">
|
||||
<spark xmlns="uri:oozie:spark-action:0.2">
|
||||
<master>yarn</master>
|
||||
<mode>cluster</mode>
|
||||
<name>Prepare association result subset of project info</name>
|
||||
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkPrepareResultProject</class>
|
||||
<jar>dhp-graph-mapper-${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}</arg>
|
||||
<arg>--outputPath</arg><arg>${workingDir}/preparedInfo</arg>
|
||||
</spark>
|
||||
<ok to="fork_extendWithProject"/>
|
||||
<error to="Kill"/>
|
||||
</action>
|
||||
|
||||
<fork name="fork_extendWithProject">
|
||||
<path start="extend_publication"/>
|
||||
<path start="extend_dataset"/>
|
||||
<path start="extend_orp"/>
|
||||
<path start="extend_software"/>
|
||||
</fork>
|
||||
|
||||
<action name="extend_publication">
|
||||
<spark xmlns="uri:oozie:spark-action:0.2">
|
||||
<master>yarn</master>
|
||||
<mode>cluster</mode>
|
||||
<name>Extend dumped publications with information about project</name>
|
||||
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkUpdateProjectInfo</class>
|
||||
<jar>dhp-graph-mapper-${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}/dump/publication</arg>
|
||||
<arg>--outputPath</arg><arg>${outputPath}/ext/publication</arg>
|
||||
<arg>--preparedInfoPath</arg><arg>${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>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>${workingDir}/dump/dataset</arg>
|
||||
<arg>--outputPath</arg><arg>${outputPath}/ext/dataset</arg>
|
||||
<arg>--preparedInfoPath</arg><arg>${workingDir}/preparedInfo</arg>
|
||||
</spark>
|
||||
<ok to="join_extend"/>
|
||||
<error to="Kill"/>
|
||||
</action>
|
||||
|
||||
<action name="extend_orp">
|
||||
<spark xmlns="uri:oozie:spark-action:0.2">
|
||||
<master>yarn</master>
|
||||
<mode>cluster</mode>
|
||||
<name>Extend dumped ORP with information about project</name>
|
||||
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkUpdateProjectInfo</class>
|
||||
<jar>dhp-graph-mapper-${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}/dump/otherresearchproduct</arg>
|
||||
<arg>--outputPath</arg><arg>${outputPath}/ext/orp</arg>
|
||||
<arg>--preparedInfoPath</arg><arg>${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>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>${workingDir}/dump/software</arg>
|
||||
<arg>--outputPath</arg><arg>${outputPath}/ext/software</arg>
|
||||
<arg>--preparedInfoPath</arg><arg>${workingDir}/preparedInfo</arg>
|
||||
</spark>
|
||||
<ok to="join_extend"/>
|
||||
<error to="Kill"/>
|
||||
</action>
|
||||
<join name="join_extend" to="End"/>
|
||||
|
||||
<end name="End"/>
|
||||
|
||||
</workflow-app>
|
||||
|
||||
|
|
@ -1,2 +0,0 @@
|
|||
## This is a classpath-based import file (this header is required)
|
||||
dump_common classpath eu/dnetlib/dhp/oa/graph/dump/wf/subworkflows/commoncommunityfunder/oozie_app
|
|
@ -77,42 +77,259 @@
|
|||
</configuration>
|
||||
</global>
|
||||
|
||||
<start to="common_action_community_funder"/>
|
||||
<start to="fork_dump"/>
|
||||
|
||||
<kill name="Kill">
|
||||
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
|
||||
</kill>
|
||||
|
||||
<action name="common_action_community_funder">
|
||||
<sub-workflow>
|
||||
<app-path>${wf:appPath()}/dump_common
|
||||
</app-path>
|
||||
<propagate-configuration/>
|
||||
<configuration>
|
||||
<property>
|
||||
<name>sourcePath</name>
|
||||
<value>${sourcePath}</value>
|
||||
</property>
|
||||
<property>
|
||||
<name>selectedResults</name>
|
||||
<value>${sourcePath}</value>
|
||||
</property>
|
||||
<property>
|
||||
<name>communityMapPath</name>
|
||||
<value>${workingDir}/communityMap</value>
|
||||
</property>
|
||||
<property>
|
||||
<name>outputPath</name>
|
||||
<value>${workingDir}</value>
|
||||
</property>
|
||||
</configuration>
|
||||
</sub-workflow>
|
||||
<ok to="splitForCommunities" />
|
||||
<error to="Kill" />
|
||||
<fork name="fork_dump">
|
||||
<path start="dump_publication"/>
|
||||
<path start="dump_dataset"/>
|
||||
<path start="dump_orp"/>
|
||||
<path start="dump_software"/>
|
||||
</fork>
|
||||
|
||||
<action name="dump_publication">
|
||||
<spark xmlns="uri:oozie:spark-action:0.2">
|
||||
<master>yarn</master>
|
||||
<mode>cluster</mode>
|
||||
<name>Dump table publication for community/funder related products</name>
|
||||
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkDumpCommunityProducts</class>
|
||||
<jar>dhp-graph-mapper-${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=${sparkSqlWarehouseDir}
|
||||
</spark-opts>
|
||||
<arg>--sourcePath</arg><arg>${selectedResults}/publication</arg>
|
||||
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Publication</arg>
|
||||
<arg>--outputPath</arg><arg>${workingDir}/dump/publication</arg>
|
||||
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
|
||||
<arg>--dumpType</arg><arg>${dumpType}</arg>
|
||||
</spark>
|
||||
<ok to="join_dump"/>
|
||||
<error to="Kill"/>
|
||||
</action>
|
||||
|
||||
<action name="dump_dataset">
|
||||
<spark xmlns="uri:oozie:spark-action:0.2">
|
||||
<master>yarn</master>
|
||||
<mode>cluster</mode>
|
||||
<name>Dump table dataset for community/funder related products</name>
|
||||
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkDumpCommunityProducts</class>
|
||||
<jar>dhp-graph-mapper-${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=${sparkSqlWarehouseDir}
|
||||
</spark-opts>
|
||||
<arg>--sourcePath</arg><arg>${selectedResults}/dataset</arg>
|
||||
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Dataset</arg>
|
||||
<arg>--outputPath</arg><arg>${workingDir}/dump/dataset</arg>
|
||||
<arg>--communityMapPath</arg><arg>${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 for community related products</name>
|
||||
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkDumpCommunityProducts</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>${selectedResults}/otherresearchproduct</arg>
|
||||
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.OtherResearchProduct</arg>
|
||||
<arg>--outputPath</arg><arg>${workingDir}/dump/otherresearchproduct</arg>
|
||||
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
|
||||
</spark>
|
||||
<ok to="join_dump"/>
|
||||
<error to="Kill"/>
|
||||
</action>
|
||||
|
||||
<action name="dump_software">
|
||||
<spark xmlns="uri:oozie:spark-action:0.2">
|
||||
<master>yarn</master>
|
||||
<mode>cluster</mode>
|
||||
<name>Dump table software for community related products</name>
|
||||
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkDumpCommunityProducts</class>
|
||||
<jar>dhp-graph-mapper-${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=${sparkSqlWarehouseDir}
|
||||
</spark-opts>
|
||||
<arg>--sourcePath</arg><arg>${selectedResults}/software</arg>
|
||||
<arg>--resultTableName</arg><arg>eu.dnetlib.dhp.schema.oaf.Software</arg>
|
||||
<arg>--outputPath</arg><arg>${workingDir}/dump/software</arg>
|
||||
<arg>--communityMapPath</arg><arg>${communityMapPath}</arg>
|
||||
</spark>
|
||||
<ok to="join_dump"/>
|
||||
<error to="Kill"/>
|
||||
</action>
|
||||
|
||||
<join name="join_dump" 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>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>${sourcePath}</arg>
|
||||
<arg>--outputPath</arg><arg>${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>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>${workingDir}/dump/publication</arg>
|
||||
<arg>--outputPath</arg><arg>${outputPath}/ext/publication</arg>
|
||||
<arg>--preparedInfoPath</arg><arg>${workingDir}/preparedInfo</arg>
|
||||
</spark>
|
||||
<ok to="join_extend"/>
|
||||
<error to="Kill"/>
|
||||
</action>
|
||||
|
||||
<action name="extend_dataset">
|
||||
<spark xmlns="uri:oozie:spark-action:0.2">
|
||||
<master>yarn</master>
|
||||
<mode>cluster</mode>
|
||||
<name>Extend dumped dataset with information about project</name>
|
||||
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkUpdateProjectInfo</class>
|
||||
<jar>dhp-graph-mapper-${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}/dump/dataset</arg>
|
||||
<arg>--outputPath</arg><arg>${outputPath}/ext/dataset</arg>
|
||||
<arg>--preparedInfoPath</arg><arg>${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>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>${workingDir}/dump/otherresearchproduct</arg>
|
||||
<arg>--outputPath</arg><arg>${outputPath}/ext/orp</arg>
|
||||
<arg>--preparedInfoPath</arg><arg>${workingDir}/preparedInfo</arg>
|
||||
</spark>
|
||||
<ok to="join_extend"/>
|
||||
<error to="Kill"/>
|
||||
</action>
|
||||
|
||||
<action name="extend_software">
|
||||
<spark xmlns="uri:oozie:spark-action:0.2">
|
||||
<master>yarn</master>
|
||||
<mode>cluster</mode>
|
||||
<name>Extend dumped software with information about project</name>
|
||||
<class>eu.dnetlib.dhp.oa.graph.dump.community.SparkUpdateProjectInfo</class>
|
||||
<jar>dhp-graph-mapper-${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}/dump/software</arg>
|
||||
<arg>--outputPath</arg><arg>${outputPath}/ext/software</arg>
|
||||
<arg>--preparedInfoPath</arg><arg>${workingDir}/preparedInfo</arg>
|
||||
</spark>
|
||||
<ok to="join_extend"/>
|
||||
<error to="Kill"/>
|
||||
</action>
|
||||
|
||||
|
||||
|
||||
<join name="join_extend" to="splitForCommunities"/>
|
||||
<action name="splitForCommunities">
|
||||
<spark xmlns="uri:oozie:spark-action:0.2">
|
||||
<master>yarn</master>
|
||||
|
|
|
@ -1,2 +0,0 @@
|
|||
## This is a classpath-based import file (this header is required)
|
||||
dump_common classpath eu/dnetlib/dhp/oa/graph/dump/wf/subworkflows/commoncommunityfunder/oozie_app
|
Loading…
Reference in New Issue