forked from D-Net/dnet-hadoop
76 lines
2.7 KiB
XML
76 lines
2.7 KiB
XML
<workflow-app name="Transformation_Workflow" xmlns="uri:oozie:workflow:0.5">
|
|
<parameters>
|
|
<property>
|
|
<name>mdstoreInputPath</name>
|
|
<description>the path of the input MDStore</description>
|
|
</property>
|
|
|
|
<property>
|
|
<name>mdstoreOutputPath</name>
|
|
<description>the path of the cleaned mdstore</description>
|
|
</property>
|
|
|
|
<property>
|
|
<name>transformationRule</name>
|
|
<description>The transformation Rule to apply</description>
|
|
</property>
|
|
|
|
<property>
|
|
<name>timestamp</name>
|
|
<description>The timestamp of the collection date</description>
|
|
</property>
|
|
|
|
<property>
|
|
<name>workflowId</name>
|
|
<description>The identifier of the workflow</description>
|
|
</property>
|
|
</parameters>
|
|
|
|
<start to="DeletePathIfExists"/>
|
|
<kill name="Kill">
|
|
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
|
|
</kill>
|
|
<action name="DeletePathIfExists">
|
|
<fs>
|
|
<mkdir path='${mdstoreOutputPath}'/>
|
|
<delete path='${mdstoreOutputPath}'/>
|
|
</fs>
|
|
<ok to="TransformJob"/>
|
|
<error to="Kill"/>
|
|
</action>
|
|
<action name="TransformJob">
|
|
<spark xmlns="uri:oozie:spark-action:0.2">
|
|
<job-tracker>${jobTracker}</job-tracker>
|
|
<name-node>${nameNode}</name-node>
|
|
<master>yarn</master>
|
|
<mode>cluster</mode>
|
|
<name>MDBuilder</name>
|
|
<class>eu.dnetlib.dhp.transformation.TransformSparkJobNode</class>
|
|
<jar>dhp-aggregations-1.0.0-SNAPSHOT.jar</jar>
|
|
<spark-opts>--num-executors 50 --conf spark.yarn.jars="hdfs://hadoop-rm1.garr-pa1.d4science.org:8020/user/oozie/share/lib/lib_20180405103059/spark2"</spark-opts>
|
|
<arg>--dateOfCollection</arg> <arg>${timestamp}</arg>
|
|
<arg>-mt</arg> <arg>yarn</arg>
|
|
<arg>--input</arg><arg>${mdstoreInputPath}</arg>
|
|
<arg>--output</arg><arg>${mdstoreOutputPath}</arg>
|
|
<arg>-w</arg><arg>${workflowId}</arg>
|
|
<arg>-tr</arg><arg>${transformationRule}</arg>
|
|
<arg>-ru</arg><arg>${rmq_user}</arg>
|
|
<arg>-rp</arg><arg>${rmq_pwd}</arg>
|
|
<arg>-rh</arg><arg>${rmq_host}</arg>
|
|
<arg>-ro</arg><arg>${rmq_ongoing}</arg>
|
|
<arg>-rr</arg><arg>${rmq_report}</arg>
|
|
</spark>
|
|
<ok to="End"/>
|
|
<error to="DropInvalidStore"/>
|
|
</action>
|
|
|
|
<action name="DropInvalidStore">
|
|
<fs>
|
|
<delete path='${mdstoreOutputPath}/../'/>
|
|
</fs>
|
|
<ok to="Kill"/>
|
|
<error to="Kill"/>
|
|
</action>
|
|
|
|
<end name="End"/>
|
|
</workflow-app> |