dnet-hadoop/dhp-workflows/dhp-aggregation/src/main/resources/eu/dnetlib/dhp/migration/wfs/regular_all_steps/oozie_app/workflow.xml

197 lines
7.4 KiB
XML

<workflow-app name="import regular entities as Graph (all steps)" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>workingPath</name>
<value>/tmp/dhp_migration</value>
<description>the base path to store temporary intermediate data</description>
</property>
<property>
<name>graphBasePath</name>
<description>the target path to store raw graph</description>
</property>
<property>
<name>reuseContent</name>
<value>false</value>
<description>should import content from the aggregator or reuse a previous version</description>
</property>
<property>
<name>postgresURL</name>
<description>the postgres URL to access to the database</description>
</property>
<property>
<name>postgresUser</name>
<description>the user postgres</description>
</property>
<property>
<name>postgresPassword</name>
<description>the password postgres</description>
</property>
<property>
<name>mongoURL</name>
<description>mongoDB url, example: mongodb://[username:password@]host[:port]</description>
</property>
<property>
<name>mongoDb</name>
<description>mongo database</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>
</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>
</configuration>
</global>
<start to="ReuseContent"/>
<kill name="Kill">
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
</kill>
<decision name="ReuseContent">
<switch>
<case to="ResetWorkingPath">${wf:conf('reuseContent') eq false}</case>
<case to="ResetAllEntitiesPath">${wf:conf('reuseContent') eq true}</case>
<default to="ResetWorkingPath"/>
</switch>
</decision>
<action name="ResetWorkingPath">
<fs>
<delete path="${workingPath}"/>
<mkdir path="${workingPath}"/>
</fs>
<ok to="ImportDB"/>
<error to="Kill"/>
</action>
<action name="ImportDB">
<java>
<main-class>eu.dnetlib.dhp.migration.step1.MigrateDbEntitiesApplication</main-class>
<arg>-p</arg><arg>${workingPath}/db_records</arg>
<arg>-pgurl</arg><arg>${postgresURL}</arg>
<arg>-pguser</arg><arg>${postgresUser}</arg>
<arg>-pgpasswd</arg><arg>${postgresPassword}</arg>
</java>
<ok to="ImportODF"/>
<error to="Kill"/>
</action>
<action name="ImportODF">
<java>
<main-class>eu.dnetlib.dhp.migration.step1.MigrateMongoMdstoresApplication</main-class>
<arg>-p</arg><arg>${workingPath}/odf_records</arg>
<arg>-mongourl</arg><arg>${mongoURL}</arg>
<arg>-mongodb</arg><arg>${mongoDb}</arg>
<arg>-f</arg><arg>ODF</arg>
<arg>-l</arg><arg>store</arg>
<arg>-i</arg><arg>cleaned</arg>
</java>
<ok to="ImportOAF"/>
<error to="Kill"/>
</action>
<action name="ImportOAF">
<java>
<main-class>eu.dnetlib.dhp.migration.step1.MigrateMongoMdstoresApplication</main-class>
<arg>-p</arg><arg>${workingPath}/oaf_records</arg>
<arg>-mongourl</arg><arg>${mongoURL}</arg>
<arg>-mongodb</arg><arg>${mongoDb}</arg>
<arg>-f</arg><arg>OAF</arg>
<arg>-l</arg><arg>store</arg>
<arg>-i</arg><arg>cleaned</arg>
</java>
<ok to="ResetAllEntitiesPath"/>
<error to="Kill"/>
</action>
<action name="ResetAllEntitiesPath">
<fs>
<delete path="${workingPath}/all_entities"/>
</fs>
<ok to="GenerateEntities"/>
<error to="Kill"/>
</action>
<action name="GenerateEntities">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>GenerateEntities</name>
<class>eu.dnetlib.dhp.migration.step2.GenerateEntitiesApplication</class>
<jar>dhp-aggregation-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory ${sparkExecutorMemory}
--executor-cores ${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.extraListeners="com.cloudera.spark.lineage.NavigatorAppListener"
--conf spark.sql.queryExecutionListeners="com.cloudera.spark.lineage.NavigatorQueryListener"
--conf spark.sql.warehouse.dir="/user/hive/warehouse"
</spark-opts>
<arg>-mt</arg> <arg>yarn-cluster</arg>
<arg>-s</arg><arg>${workingPath}/db_records,${workingPath}/oaf_records,${workingPath}/odf_records</arg>
<arg>-t</arg><arg>${workingPath}/all_entities</arg>
<arg>-pgurl</arg><arg>${postgresURL}</arg>
<arg>-pguser</arg><arg>${postgresUser}</arg>
<arg>-pgpasswd</arg><arg>${postgresPassword}</arg>
</spark>
<ok to="ResetGraphPath"/>
<error to="Kill"/>
</action>
<action name="ResetGraphPath">
<fs>
<delete path="${graphBasePath}/graph_raw"/>
<mkdir path="${graphBasePath}/graph_raw"/>
</fs>
<ok to="GenerateGraph"/>
<error to="Kill"/>
</action>
<action name="GenerateGraph">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>GenerateGraph</name>
<class>eu.dnetlib.dhp.migration.step3.DispatchEntitiesApplication</class>
<jar>dhp-aggregation-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory ${sparkExecutorMemory}
--executor-cores ${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.extraListeners="com.cloudera.spark.lineage.NavigatorAppListener"
--conf spark.sql.queryExecutionListeners="com.cloudera.spark.lineage.NavigatorQueryListener"
--conf spark.sql.warehouse.dir="/user/hive/warehouse"
</spark-opts>
<arg>-mt</arg> <arg>yarn-cluster</arg>
<arg>-s</arg><arg>${workingPath}/all_entities</arg>
<arg>-g</arg><arg>${graphBasePath}/graph_raw</arg>
</spark>
<ok to="End"/>
<error to="Kill"/>
</action>
<end name="End"/>
</workflow-app>