read from blacklist and write the result as relations on hdfs

This commit is contained in:
Miriam Baglioni 2020-04-29 18:16:01 +02:00
parent 869f576273
commit 6a47e6191d
3 changed files with 37 additions and 213 deletions

View File

@ -10,6 +10,24 @@
<modelVersion>4.0.0</modelVersion>
<artifactId>dhp-blacklist</artifactId>
<dependencies>
<dependency>
<groupId>eu.dnetlib.dhp</groupId>
<artifactId>dhp-graph-mapper</artifactId>
<version>1.1.7-SNAPSHOT</version>
<scope>compile</scope>
</dependency>
<dependency>
<groupId>com.fasterxml.jackson.core</groupId>
<artifactId>jackson-databind</artifactId>
<scope>compile</scope>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-common</artifactId>
<scope>compile</scope>
</dependency>
</dependencies>
</project>

View File

@ -5,6 +5,12 @@
"paramDescription": "the path where storing the sequential file",
"paramRequired": true
},
{
"paramName": "nn",
"paramLongName": "hdfsNameNode",
"paramDescription": "the name node on hdfs",
"paramRequired": true
},
{
"paramName": "pgurl",
"paramLongName": "postgresUrl",
@ -22,11 +28,5 @@
"paramLongName": "postgresPassword",
"paramDescription": "postgres password",
"paramRequired": false
},
{
"paramName": "a",
"paramLongName": "action",
"paramDescription": "process claims",
"paramRequired": false
}
]

View File

@ -1,34 +1,4 @@
<workflow-app name="bulk_tagging" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>sourcePath</name>
<description>the source path</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>isLookUpUrl</name>
<description>the isLookup service endpoint</description>
</property>
<property>
<name>pathMap</name>
<description>the json path associated to each selection field</description>
</property>
<property>
<name>outputPath</name>
<description>the output path</description>
</property>
</parameters>
<workflow-app name="blacklsting" xmlns="uri:oozie:workflow:0.5">
<start to="reset-outputpath"/>
@ -38,190 +8,26 @@
<action name="reset-outputpath">
<fs>
<delete path='${outputPath}/relation'/>
<delete path='${outputPath}/publication'/>
<delete path='${outputPath}/dataset'/>
<delete path='${outputPath}/otherresearchproduct'/>
<delete path='${outputPath}/software'/>
<delete path='${outputPath}/organization'/>
<delete path='${outputPath}/project'/>
<delete path='${outputPath}/datasource'/>
<delete path='${hdfsPath}'/>
</fs>
<ok to="copy_entities"/>
<ok to="read_blacklist"/>
<error to="Kill"/>
</action>
<fork name="copy_entities">
<path start="copy_relation"/>
<path start="copy_organization"/>
<path start="copy_projects"/>
<path start="copy_datasources"/>
</fork>
<action name="copy_relation">
<distcp xmlns="uri:oozie:distcp-action:0.2">
<action name="read_blacklist">
<java>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<arg>${nameNode}/${sourcePath}/relation</arg>
<arg>${nameNode}/${outputPath}/relation</arg>
</distcp>
<ok to="copy_wait"/>
<main-class>eu.dnetlib.dhp.blacklist.ReadBlacklistFromDB</main-class>
<arg>--hdfsPath</arg><arg>${workingDir}/blacklist</arg>
<arg>--hdfsNameNode</arg><arg>${nameNode}</arg>
<arg>--postgresUrl</arg><arg>${postgresUrl}</arg>
<arg>--postgresUser</arg><arg>${postgresUser}</arg>
<arg>--postgresPassword</arg><arg>${postgresPassword}</arg>
</java>
<ok to="End"/>
<error to="Kill"/>
</action>
<action name="copy_organization">
<distcp xmlns="uri:oozie:distcp-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<arg>${nameNode}/${sourcePath}/organization</arg>
<arg>${nameNode}/${outputPath}/organization</arg>
</distcp>
<ok to="copy_wait"/>
<error to="Kill"/>
</action>
<action name="copy_projects">
<distcp xmlns="uri:oozie:distcp-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<arg>${nameNode}/${sourcePath}/project</arg>
<arg>${nameNode}/${outputPath}/project</arg>
</distcp>
<ok to="copy_wait"/>
<error to="Kill"/>
</action>
<action name="copy_datasources">
<distcp xmlns="uri:oozie:distcp-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<arg>${nameNode}/${sourcePath}/datasource</arg>
<arg>${nameNode}/${outputPath}/datasource</arg>
</distcp>
<ok to="copy_wait"/>
<error to="Kill"/>
</action>
<join name="copy_wait" to="fork_exec_bulktag"/>
<fork name="fork_exec_bulktag">
<path start="join_bulktag_publication"/>
<path start="join_bulktag_dataset"/>
<path start="join_bulktag_otherresearchproduct"/>
<path start="join_bulktag_software"/>
</fork>
<action name="join_bulktag_publication">
<spark xmlns="uri:oozie:spark-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>bulkTagging-publication</name>
<class>eu.dnetlib.dhp.bulktag.SparkBulkTagJob2</class>
<jar>dhp-bulktag-${projectVersion}.jar</jar>
<spark-opts>
--num-executors=${sparkExecutorNumber}
--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}
</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}/publication</arg>
<arg>--pathMap</arg><arg>${pathMap}</arg>
<arg>--isLookUpUrl</arg><arg>${isLookUpUrl}</arg>
</spark>
<ok to="wait"/>
<error to="Kill"/>
</action>
<action name="join_bulktag_dataset">
<spark xmlns="uri:oozie:spark-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>bulkTagging-dataset</name>
<class>eu.dnetlib.dhp.bulktag.SparkBulkTagJob2</class>
<jar>dhp-bulktag-${projectVersion}.jar</jar>
<spark-opts>
--num-executors=${sparkExecutorNumber}
--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}
</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}/dataset</arg>
<arg>--pathMap</arg><arg>${pathMap}</arg>
<arg>--isLookUpUrl</arg><arg>${isLookUpUrl}</arg>
</spark>
<ok to="wait"/>
<error to="Kill"/>
</action>
<action name="join_bulktag_otherresearchproduct">
<spark xmlns="uri:oozie:spark-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>bulkTagging-orp</name>
<class>eu.dnetlib.dhp.bulktag.SparkBulkTagJob2</class>
<jar>dhp-bulktag-${projectVersion}.jar</jar>
<spark-opts>
--num-executors=${sparkExecutorNumber}
--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}
</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}/otherresearchproduct</arg>
<arg>--pathMap</arg><arg>${pathMap}</arg>
<arg>--isLookUpUrl</arg><arg>${isLookUpUrl}</arg>
</spark>
<ok to="wait"/>
<error to="Kill"/>
</action>
<action name="join_bulktag_software">
<spark xmlns="uri:oozie:spark-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>bulkTagging-software</name>
<class>eu.dnetlib.dhp.bulktag.SparkBulkTagJob2</class>
<jar>dhp-bulktag-${projectVersion}.jar</jar>
<spark-opts>
--num-executors=${sparkExecutorNumber}
--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}
</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}/software</arg>
<arg>--pathMap</arg><arg>${pathMap}</arg>
<arg>--isLookUpUrl</arg><arg>${isLookUpUrl}</arg>
</spark>
<ok to="wait"/>
<error to="Kill"/>
</action>
<join name="wait" to="End"/>
<end name="End"/>
</workflow-app>