defined a single oozie workflow for the generation of doiboost

This commit is contained in:
Sandro La Bruzzo 2021-01-04 17:01:35 +01:00
parent 7185158942
commit b0dc92786f
4 changed files with 372 additions and 2 deletions

View File

@ -39,6 +39,7 @@ object SparkGenerateDOIBoostActionSet {
val dbaffiliationRelationPath = parser.get("dbaffiliationRelationPath") val dbaffiliationRelationPath = parser.get("dbaffiliationRelationPath")
val dbOrganizationPath = parser.get("dbOrganizationPath") val dbOrganizationPath = parser.get("dbOrganizationPath")
val workingDirPath = parser.get("targetPath") val workingDirPath = parser.get("targetPath")
val sequenceFilePath = parser.get("sFilePath")
spark.read.load(dbDatasetPath).as[OafDataset] spark.read.load(dbDatasetPath).as[OafDataset]
.map(d =>DoiBoostMappingUtil.fixResult(d)) .map(d =>DoiBoostMappingUtil.fixResult(d))
@ -65,7 +66,7 @@ object SparkGenerateDOIBoostActionSet {
val d: Dataset[(String, String)] =spark.read.load(s"$workingDirPath/actionSet").as[(String,String)] val d: Dataset[(String, String)] =spark.read.load(s"$workingDirPath/actionSet").as[(String,String)]
d.rdd.map(s => (new Text(s._1), new Text(s._2))).saveAsHadoopFile(s"$workingDirPath/rawset", classOf[Text], classOf[Text], classOf[SequenceFileOutputFormat[Text,Text]], classOf[GzipCodec]) d.rdd.repartition(6000).map(s => (new Text(s._1), new Text(s._2))).saveAsHadoopFile(s"$sequenceFilePath", classOf[Text], classOf[Text], classOf[SequenceFileOutputFormat[Text,Text]], classOf[GzipCodec])

View File

@ -5,5 +5,6 @@
{"paramName": "cr", "paramLongName":"crossRefRelation", "paramDescription": "the UnpayWall Publication Path", "paramRequired": true}, {"paramName": "cr", "paramLongName":"crossRefRelation", "paramDescription": "the UnpayWall Publication Path", "paramRequired": true},
{"paramName": "da", "paramLongName":"dbaffiliationRelationPath", "paramDescription": "the MAG Publication Path", "paramRequired": true}, {"paramName": "da", "paramLongName":"dbaffiliationRelationPath", "paramDescription": "the MAG Publication Path", "paramRequired": true},
{"paramName": "do", "paramLongName":"dbOrganizationPath", "paramDescription": "the MAG Publication Path", "paramRequired": true}, {"paramName": "do", "paramLongName":"dbOrganizationPath", "paramDescription": "the MAG Publication Path", "paramRequired": true},
{"paramName": "w", "paramLongName":"targetPath", "paramDescription": "the Working Path", "paramRequired": true} {"paramName": "w", "paramLongName":"targetPath", "paramDescription": "the Working Path", "paramRequired": true},
{"paramName": "sp", "paramLongName":"sFilePath", "paramDescription": "the Sequence file Path", "paramRequired": true},
] ]

View File

@ -0,0 +1,42 @@
<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>oozie.action.sharelib.for.spark</name>
<value>spark2</value>
</property>
<property>
<name>oozie.launcher.mapreduce.user.classpath.first</name>
<value>true</value>
</property>
<property>
<name>hive_metastore_uris</name>
<value>thrift://iis-cdh5-test-m3.ocean.icm.edu.pl:9083</value>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<value>http://iis-cdh5-test-gw.ocean.icm.edu.pl:18089</value>
</property>
<property>
<name>spark2EventLogDir</name>
<value>/user/spark/spark2ApplicationHistory</value>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>"com.cloudera.spark.lineage.NavigatorAppListener"</value>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>"com.cloudera.spark.lineage.NavigatorQueryListener"</value>
</property>
</configuration>

View File

@ -0,0 +1,326 @@
<workflow-app name="Generate DOIBoost ActionSet" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>sparkDriverMemory</name>
<description>memory for driver process</description>
</property>
<property>
<name>sparkExecutorMemory</name>
<description>memory for individual executor</description>
</property>
<property>
<name>sparkExecutorIntersectionMemory</name>
<description>memory for individual executor</description>
</property>
<property>
<name>sparkExecutorCores</name>
<description>number of cores used by single executor</description>
</property>
<!-- Itersection Parameters -->
<property>
<name>workingPath</name>
<description>the working Path</description>
</property>
<property>
<name>hostedByMapPath</name>
<description>the hostedByMap Path</description>
</property>
<property>
<name>outputPath</name>
<description>the Path of the sequence file action set</description>
</property>
<!-- Crossref Parameters -->
<property>
<name>inputPathCrossref</name>
<description>the Crossref input path</description>
</property>
<property>
<name>crossrefTimestamp</name>
<description>Timestamp for the Crossref incremental Harvesting</description>
</property>
<!-- MAG Parameters -->
<property>
<name>inputPathMAG</name>
<description>the MAG working path</description>
</property>
<!-- UnpayWall Parameters -->
<property>
<name>inputPathUnpayWall</name>
<description>the UnpayWall working path</description>
</property>
<!-- ORCID Parameters -->
<property>
<name>inputPathOrcid</name>
<description>the ORCID working path</description>
</property>
</parameters>
<start to="resume_from"/>
<decision name="resume_from">
<switch>
<case to="ConvertCrossrefToOAF">${wf:conf('resumeFrom') eq 'ConvertCrossrefToOAF'}</case>
<case to="ResetMagWorkingPath">${wf:conf('resumeFrom') eq 'ResetMagWorkingPath'}</case>
<case to="PreprocessMag">${wf:conf('resumeFrom') eq 'PreprocessMag'}</case>
<case to="PreprocessUW">${wf:conf('resumeFrom') eq 'PreprocessUW'}</case>
<case to="PreprocessORCID">${wf:conf('resumeFrom') eq 'PreprocessORCID'}</case>
<case to="CreateDOIBoost">${wf:conf('resumeFrom') eq 'CreateDOIBoost'}</case>
<case to="GenerateActionSet">${wf:conf('resumeFrom') eq 'GenerateActionSet'}</case>
<default to="ImportCrossRef"/>
</switch>
</decision>
<kill name="Kill">
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
</kill>
<action name="ImportCrossRef">
<java>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<main-class>eu.dnetlib.doiboost.crossref.CrossrefImporter</main-class>
<arg>-t</arg><arg>${inputPathCrossref}/index_update</arg>
<arg>-n</arg><arg>${nameNode}</arg>
<arg>-ts</arg><arg>${timestamp}</arg>
</java>
<ok to="GenerateCrossrefDataset"/>
<error to="Kill"/>
</action>
<!-- CROSSREF SECTION -->
<action name="GenerateCrossrefDataset">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>GenerateCrossrefDataset</name>
<class>eu.dnetlib.doiboost.crossref.CrossrefDataset</class>
<jar>dhp-doiboost-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.sql.shuffle.partitions=3840
${sparkExtraOPT}
</spark-opts>
<arg>--workingPath</arg><arg>${inputPathCrossref}</arg>
<arg>--master</arg><arg>yarn-cluster</arg>
</spark>
<ok to="RenameDataset"/>
<error to="Kill"/>
</action>
<action name="RenameDataset">
<fs>
<delete path="${inputPathCrossref}/crossref_ds"/>
<move source="${inputPathCrossref}/crossref_ds_updated"
target="${inputPathCrossref}/crossref_ds"/>
</fs>
<ok to="ConvertCrossrefToOAF"/>
<error to="Kill"/>
</action>
<action name="ConvertCrossrefToOAF">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>ConvertCrossrefToOAF</name>
<class>eu.dnetlib.doiboost.crossref.SparkMapDumpIntoOAF</class>
<jar>dhp-doiboost-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.sql.shuffle.partitions=3840
${sparkExtraOPT}
</spark-opts>
<arg>--sourcePath</arg><arg>${inputPathCrossref}/crossref_ds</arg>
<arg>--targetPath</arg><arg>${workingPath}</arg>
<arg>--master</arg><arg>yarn-cluster</arg>
</spark>
<ok to="ResetMagWorkingPath"/>
<error to="Kill"/>
</action>
<!-- MAG SECTION -->
<action name="ResetMagWorkingPath">
<fs>
<delete path='${inputPathMAG}/dataset'/>
<delete path='${inputPathMAG}/process'/>
<delete path='${inputPathMAG}/dataset'/>
</fs>
<ok to="ConvertMagToDataset"/>
<error to="Kill"/>
</action>
<action name="ConvertMagToDataset">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>Convert Mag to Dataset</name>
<class>eu.dnetlib.doiboost.mag.SparkImportMagIntoDataset</class>
<jar>dhp-doiboost-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
${sparkExtraOPT}
</spark-opts>
<arg>--sourcePath</arg><arg>${inputPathMAG}/input</arg>
<arg>--targetPath</arg><arg>${inputPathMAG}/dataset</arg>
<arg>--master</arg><arg>yarn-cluster</arg>
</spark>
<ok to="PreprocessMag"/>
<error to="Kill"/>
</action>
<action name="PreprocessMag">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>Convert Mag to OAF Dataset</name>
<class>eu.dnetlib.doiboost.mag.SparkPreProcessMAG</class>
<jar>dhp-doiboost-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.sql.shuffle.partitions=3840
${sparkExtraOPT}
</spark-opts>
<arg>--sourcePath</arg><arg>${inputPathMAG}/dataset</arg>
<arg>--workingPath</arg><arg>${inputPathMAG}/process</arg>
<arg>--targetPath</arg><arg>${workingPath}</arg>
<arg>--master</arg><arg>yarn-cluster</arg>
</spark>
<ok to="PreprocessUW"/>
<error to="Kill"/>
</action>
<!-- UnpayWall SECTION -->
<action name="PreprocessUW">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>Convert UnpayWall to Dataset</name>
<class>eu.dnetlib.doiboost.uw.SparkMapUnpayWallToOAF</class>
<jar>dhp-doiboost-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.sql.shuffle.partitions=3840
${sparkExtraOPT}
</spark-opts>
<arg>--sourcePath</arg><arg>${inputPathUnpayWall}/uw_extracted</arg>
<arg>--targetPath</arg><arg>${workingPath}</arg>
<arg>--master</arg><arg>yarn-cluster</arg>
</spark>
<ok to="PreprocessORCID"/>
<error to="Kill"/>
</action>
<!-- ORCID SECTION -->
<action name="PreprocessORCID">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>Convert ORCID to Dataset</name>
<class>eu.dnetlib.doiboost.orcid.SparkConvertORCIDToOAF</class>
<jar>dhp-doiboost-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.sql.shuffle.partitions=3840
${sparkExtraOPT}
</spark-opts>
<arg>--sourcePath</arg><arg>${inputPathOrcid}</arg>
<arg>--targetPath</arg><arg>${workingPath}</arg>
<arg>--master</arg><arg>yarn-cluster</arg>
</spark>
<ok to="CreateDOIBoost"/>
<error to="Kill"/>
</action>
<!-- INTERSECTION SECTION-->
<action name="CreateDOIBoost">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>Create DOIBoost Infospace</name>
<class>eu.dnetlib.doiboost.SparkGenerateDoiBoost</class>
<jar>dhp-doiboost-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.sql.shuffle.partitions=3840
${sparkExtraOPT}
</spark-opts>
<arg>--hostedByMapPath</arg><arg>${hostedByMapPath}</arg>
<arg>--affiliationPath</arg><arg>${inputPathMAG}/process/Affiliations</arg>
<arg>--paperAffiliationPath</arg><arg>${inputPathMAG}/process/PaperAuthorAffiliations</arg>
<arg>--workingDirPath</arg><arg>${workingPath}</arg>
<arg>--master</arg><arg>yarn-cluster</arg>
</spark>
<ok to="GenerateActionSet"/>
<error to="Kill"/>
</action>
<action name="GenerateActionSet">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>Generate DOIBoost ActionSet</name>
<class>eu.dnetlib.doiboost.SparkGenerateDOIBoostActionSet</class>
<jar>dhp-doiboost-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${sparkExecutorIntersectionMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.sql.shuffle.partitions=3840
${sparkExtraOPT}
</spark-opts>
<arg>--dbPublicationPath</arg><arg>${workingPath}/doiBoostPublicationFiltered</arg>
<arg>--dbDatasetPath</arg><arg>${workingPath}/crossrefDataset</arg>
<arg>--crossRefRelation</arg><arg>${workingPath}/crossrefRelation</arg>
<arg>--dbaffiliationRelationPath</arg><arg>${workingPath}/doiBoostPublicationAffiliation</arg>
<arg>-do</arg><arg>${workingPath}/doiBoostOrganization</arg>
<arg>--targetPath</arg><arg>${workingPath}/actionDataSet</arg>
<arg>--sFilePath</arg><arg>${outputPath}</arg>
<arg>--master</arg><arg>yarn-cluster</arg>
</spark>
<ok to="End"/>
<error to="Kill"/>
</action>
<end name="End"/>
</workflow-app>