dnet-hadoop/dhp-workflows/dhp-graph-provision/src/main/resources/eu/dnetlib/dhp/sx/solr/provision/oozie_app/workflow.xml

113 lines
4.6 KiB
XML

<workflow-app name="Index Scholexplorer Infospace" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>sourcePath</name>
<description>the sourcePath of the json RDDs</description>
</property>
<property>
<name>isLookupUrl</name>
<description>URL for the isLookup service</description>
</property>
<property>
<name>solrDeletionQuery</name>
<value>*:*</value>
<description>query used in the deleted by query operation</description>
</property>
<property>
<name>format</name>
<description>metadata format name (SMF)</description>
</property>
</parameters>
<start to="indexScholix"/>
<kill name="Kill">
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
</kill>
<action name="drop_solr_collection">
<java>
<configuration>
<property>
<name>oozie.launcher.mapreduce.user.classpath.first</name>
<value>true</value>
</property>
</configuration>
<main-class>eu.dnetlib.dhp.oa.provision.SolrAdminApplication</main-class>
<arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
<arg>--format</arg><arg>${format}</arg>
<arg>--action</arg><arg>DELETE_BY_QUERY</arg>
<arg>--query</arg><arg>${solrDeletionQuery}</arg>
<arg>--commit</arg><arg>true</arg>
</java>
<ok to="create_solr_index"/>
<error to="Kill"/>
</action>
<action name="create_solr_index">
<java>
<configuration>
<property>
<name>oozie.launcher.mapreduce.user.classpath.first</name>
<value>true</value>
</property>
</configuration>
<main-class>eu.dnetlib.dhp.oa.provision.SolrAdminApplication</main-class>
<arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
<arg>--format</arg><arg>${format}</arg>
<arg>--action</arg><arg>CREATE</arg>
</java>
<ok to="indexScholix"/>
<error to="Kill"/>
</action>
<action name="indexScholix">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Index summary</name>
<class>eu.dnetlib.dhp.sx.provision.SparkIndexCollectionOnSOLR</class>
<jar>dhp-graph-provision-${projectVersion}.jar</jar>
<spark-opts>
--conf spark.dynamicAllocation.enabled=true
--conf spark.shuffle.service.enabled=true
--executor-memory=${sparkExecutorMemory}
--conf spark.dynamicAllocation.maxExecutors="16"
--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>--cluster</arg><arg>yarn</arg>
<arg>--isURL</arg><arg>${isLookupUrl}</arg>
<arg>--inputPath</arg><arg>${sourcePath}</arg>
<arg>--format</arg><arg>${format}</arg>
</spark>
<ok to="commit_solr_collection"/>
<error to="Kill"/>
</action>
<action name="commit_solr_collection">
<java>
<configuration>
<property>
<name>oozie.launcher.mapreduce.user.classpath.first</name>
<value>true</value>
</property>
</configuration>
<main-class>eu.dnetlib.dhp.oa.provision.SolrAdminApplication</main-class>
<arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
<arg>--format</arg><arg>${format}</arg>
<arg>--action</arg><arg>COMMIT</arg>
</java>
<ok to="End"/>
<error to="Kill"/>
</action>
<end name="End"/>
</workflow-app>