2020-06-29 16:33:32 +02:00
|
|
|
<workflow-app name="create broker events - partial" xmlns="uri:oozie:workflow:0.5">
|
2020-06-23 08:37:35 +02:00
|
|
|
|
|
|
|
<parameters>
|
|
|
|
<property>
|
|
|
|
<name>graphInputPath</name>
|
|
|
|
<description>the path where the graph is stored</description>
|
|
|
|
</property>
|
|
|
|
<property>
|
|
|
|
<name>workingPath</name>
|
|
|
|
<description>the path where the the generated data will be stored</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>${oozieActionShareLibForSpark2}</value>
|
|
|
|
</property>
|
|
|
|
</configuration>
|
|
|
|
</global>
|
|
|
|
|
2020-07-18 09:40:36 +02:00
|
|
|
<start to="generate_events"/>
|
2020-06-23 08:37:35 +02:00
|
|
|
|
|
|
|
<kill name="Kill">
|
|
|
|
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
|
|
|
|
</kill>
|
2020-06-30 16:17:09 +02:00
|
|
|
|
2020-07-15 09:18:40 +02:00
|
|
|
<action name="generate_events">
|
|
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
|
|
<master>yarn</master>
|
|
|
|
<mode>cluster</mode>
|
|
|
|
<name>GenerateEventsJob</name>
|
|
|
|
<class>eu.dnetlib.dhp.broker.oa.GenerateEventsJob</class>
|
|
|
|
<jar>dhp-broker-events-${projectVersion}.jar</jar>
|
|
|
|
<spark-opts>
|
|
|
|
--executor-cores=${sparkExecutorCores}
|
|
|
|
--executor-memory=${sparkExecutorMemory}
|
|
|
|
--driver-memory=${sparkDriverMemory}
|
|
|
|
--conf spark.extraListeners=${spark2ExtraListeners}
|
|
|
|
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
|
|
|
|
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
|
|
|
|
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
|
|
|
|
--conf spark.sql.shuffle.partitions=3840
|
|
|
|
</spark-opts>
|
|
|
|
<arg>--workingPath</arg><arg>${workingPath}</arg>
|
|
|
|
<arg>--datasourceIdWhitelist</arg><arg>${datasourceIdWhitelist}</arg>
|
|
|
|
<arg>--datasourceTypeWhitelist</arg><arg>${datasourceTypeWhitelist}</arg>
|
|
|
|
<arg>--datasourceIdBlacklist</arg><arg>${datasourceIdBlacklist}</arg>
|
|
|
|
</spark>
|
|
|
|
<ok to="index_es"/>
|
|
|
|
<error to="Kill"/>
|
|
|
|
</action>
|
|
|
|
|
|
|
|
<action name="index_es">
|
|
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
|
|
<master>yarn</master>
|
|
|
|
<mode>cluster</mode>
|
|
|
|
<name>IndexOnESJob</name>
|
|
|
|
<class>eu.dnetlib.dhp.broker.oa.IndexOnESJob</class>
|
|
|
|
<jar>dhp-broker-events-${projectVersion}.jar</jar>
|
|
|
|
<spark-opts>
|
|
|
|
--executor-memory=${sparkExecutorMemory}
|
|
|
|
--driver-memory=${sparkDriverMemory}
|
|
|
|
--conf spark.dynamicAllocation.maxExecutors="8"
|
|
|
|
--conf spark.extraListeners=${spark2ExtraListeners}
|
|
|
|
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
|
|
|
|
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
|
|
|
|
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
|
|
|
|
--conf spark.sql.shuffle.partitions=3840
|
|
|
|
</spark-opts>
|
|
|
|
<arg>--workingPath</arg><arg>${workingPath}</arg>
|
|
|
|
<arg>--index</arg><arg>${esIndexName}</arg>
|
|
|
|
<arg>--esHost</arg><arg>${esIndexHost}</arg>
|
|
|
|
</spark>
|
|
|
|
<ok to="stats"/>
|
|
|
|
<error to="Kill"/>
|
|
|
|
</action>
|
|
|
|
|
|
|
|
<action name="stats">
|
2020-06-23 08:37:35 +02:00
|
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
|
|
<master>yarn</master>
|
|
|
|
<mode>cluster</mode>
|
2020-07-10 16:12:08 +02:00
|
|
|
<name>GenerateStatsJob</name>
|
|
|
|
<class>eu.dnetlib.dhp.broker.oa.GenerateStatsJob</class>
|
2020-06-23 08:37:35 +02:00
|
|
|
<jar>dhp-broker-events-${projectVersion}.jar</jar>
|
|
|
|
<spark-opts>
|
2020-07-07 15:37:13 +02:00
|
|
|
--executor-cores=${sparkExecutorCores}
|
2020-06-23 08:37:35 +02:00
|
|
|
--executor-memory=${sparkExecutorMemory}
|
|
|
|
--driver-memory=${sparkDriverMemory}
|
|
|
|
--conf spark.extraListeners=${spark2ExtraListeners}
|
|
|
|
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
|
|
|
|
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
|
|
|
|
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
|
|
|
|
--conf spark.sql.shuffle.partitions=3840
|
|
|
|
</spark-opts>
|
2020-07-07 15:37:13 +02:00
|
|
|
<arg>--graphPath</arg><arg>${graphInputPath}</arg>
|
2020-06-23 08:37:35 +02:00
|
|
|
<arg>--workingPath</arg><arg>${workingPath}</arg>
|
|
|
|
</spark>
|
|
|
|
<ok to="End"/>
|
|
|
|
<error to="Kill"/>
|
|
|
|
</action>
|
|
|
|
|
|
|
|
<end name="End"/>
|
|
|
|
|
|
|
|
</workflow-app>
|