172 lines
7.0 KiB
XML
172 lines
7.0 KiB
XML
<workflow-app name="Publication Features Extraction" xmlns="uri:oozie:workflow:0.5">
|
|
<parameters>
|
|
<property>
|
|
<name>publicationsPath</name>
|
|
<description>the input entity path</description>
|
|
</property>
|
|
<property>
|
|
<name>workingPath</name>
|
|
<description>path for the working directory</description>
|
|
</property>
|
|
<property>
|
|
<name>numPartitions</name>
|
|
<description>number of partitions for the spark files</description>
|
|
</property>
|
|
<property>
|
|
<name>featuresPath</name>
|
|
<description>location of the embeddings</description>
|
|
</property>
|
|
<property>
|
|
<name>topicsPath</name>
|
|
<description>location of the topics</description>
|
|
</property>
|
|
<property>
|
|
<name>outputPath</name>
|
|
<description>location of the output authors</description>
|
|
</property>
|
|
<property>
|
|
<name>bertModel</name>
|
|
<description>location of the bert model</description>
|
|
</property>
|
|
<property>
|
|
<name>bertSentenceModel</name>
|
|
<description>location of the bert sentence model</description>
|
|
</property>
|
|
<property>
|
|
<name>wordEmbeddingsModel</name>
|
|
<description>location of the word embeddings model</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>${oozieActionShareLibForSpark2}</value>
|
|
</property>
|
|
</configuration>
|
|
</global>
|
|
|
|
<start to="resetWorkingPath"/>
|
|
|
|
<kill name="Kill">
|
|
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
|
|
</kill>
|
|
|
|
<action name="resetWorkingPath">
|
|
<fs>
|
|
<delete path="${workingPath}"/>
|
|
</fs>
|
|
<ok to="PublicationFeatureExtractor"/>
|
|
<error to="Kill"/>
|
|
</action>
|
|
|
|
<!--TODO reimplement LDA procedure and put inference here as the first step-->
|
|
|
|
<action name="PublicationFeatureExtractor">
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
<master>yarn</master>
|
|
<mode>cluster</mode>
|
|
<name>Publication Feature Extraction</name>
|
|
<class>eu.dnetlib.jobs.featureextraction.SparkPublicationFeatureExtractor</class>
|
|
<jar>dnet-and-test-${projectVersion}.jar</jar>
|
|
<spark-opts>
|
|
--num-executors=32
|
|
--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.shuffle.partitions=3840
|
|
--conf spark.dynamicAllocation.enabled=false
|
|
</spark-opts>
|
|
<arg>--publicationsPath</arg><arg>${publicationsPath}</arg>
|
|
<arg>--workingPath</arg><arg>${workingPath}</arg>
|
|
<arg>--numPartitions</arg><arg>${numPartitions}</arg>
|
|
<arg>--featuresPath</arg><arg>${featuresPath}</arg>
|
|
<arg>--wordEmbeddingsModel</arg><arg>${wordEmbeddingsModel}</arg>
|
|
<arg>--bertModel</arg><arg>${bertModel}</arg>
|
|
<arg>--bertSentenceModel</arg><arg>${bertSentenceModel}</arg>
|
|
</spark>
|
|
<ok to="AuthorExtractor"/>
|
|
<error to="Kill"/>
|
|
</action>
|
|
|
|
<action name="AuthorExtractor">
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
<master>yarn</master>
|
|
<mode>cluster</mode>
|
|
<name>Author Extraction</name>
|
|
<class>eu.dnetlib.jobs.featureextraction.SparkAuthorExtractor</class>
|
|
<jar>dnet-and-test-${projectVersion}.jar</jar>
|
|
<spark-opts>
|
|
--num-executors=32
|
|
--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.shuffle.partitions=3840
|
|
--conf spark.dynamicAllocation.enabled=true
|
|
</spark-opts>
|
|
<arg>--workingPath</arg><arg>${workingPath}</arg>
|
|
<arg>--numPartitions</arg><arg>${numPartitions}</arg>
|
|
<arg>--publicationsPath</arg><arg>${publicationsPath}</arg>
|
|
<arg>--topicsPath</arg><arg>${topicsPath}</arg>
|
|
<arg>--featuresPath</arg><arg>${featuresPath}</arg>
|
|
<arg>--outputPath</arg><arg>${outputPath}</arg>
|
|
</spark>
|
|
<ok to="End"/>
|
|
<error to="Kill"/>
|
|
</action>
|
|
|
|
<end name="End"/>
|
|
</workflow-app> |