dnet-hadoop/dhp-workflows/dhp-doiboost/src/main/resources/eu/dnetlib/dhp/doiboost/orcid_extract_xml_activities/oozie_app/workflow.xml

232 lines
8.7 KiB
XML

<workflow-app name="Extract Orcid XML Works From Activities" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>workingPath</name>
<description>the working dir base path</description>
</property>
</parameters>
<global>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<configuration>
<property>
<name>oozie.action.sharelib.for.java</name>
<value>${oozieActionShareLibForSpark2}</value>
</property>
<property>
<name>oozie.launcher.mapreduce.user.classpath.first</name>
<value>true</value>
</property>
<property>
<name>oozie.launcher.mapreduce.map.java.opts</name>
<value>-Xmx2g</value>
</property>
<property>
<name>oozie.use.system.libpath</name>
<value>true</value>
</property>
</configuration>
</global>
<start to="ResetWorkingPath"/>
<kill name="Kill">
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
</kill>
<action name="ResetWorkingPath">
<fs>
<delete path='${workingPath}/xml/works'/>
<mkdir path='${workingPath}/xml/works'/>
</fs>
<ok to="fork_node"/>
<error to="Kill"/>
</action>
<fork name = "fork_node">
<path start = "ExtractXMLWorkActivities_0"/>
<path start = "ExtractXMLWorkActivities_1"/>
<path start = "ExtractXMLWorkActivities_2"/>
<path start = "ExtractXMLWorkActivities_3"/>
<path start = "ExtractXMLWorkActivities_4"/>
<path start = "ExtractXMLWorkActivities_5"/>
<path start = "ExtractXMLWorkActivities_6"/>
<path start = "ExtractXMLWorkActivities_7"/>
<path start = "ExtractXMLWorkActivities_8"/>
<path start = "ExtractXMLWorkActivities_9"/>
<path start = "ExtractXMLWorkActivities_X"/>
</fork>
<action name="ExtractXMLWorkActivities_0">
<java>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<main-class>eu.dnetlib.doiboost.orcid.ExtractXMLActivitiesData</main-class>
<arg>-w</arg><arg>${workingPath}/</arg>
<arg>-n</arg><arg>${nameNode}</arg>
<arg>-f</arg><arg>ORCID_2020_10_activites_0.tar.gz</arg>
<arg>-ow</arg><arg>xml/works/xml_works_0.seq</arg>
<arg>-oew</arg><arg>---</arg>
</java>
<ok to="join_node"/>
<error to="Kill"/>
</action>
<action name="ExtractXMLWorkActivities_1">
<java>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<main-class>eu.dnetlib.doiboost.orcid.ExtractXMLActivitiesData</main-class>
<arg>-w</arg><arg>${workingPath}/</arg>
<arg>-n</arg><arg>${nameNode}</arg>
<arg>-f</arg><arg>ORCID_2020_10_activites_1.tar.gz</arg>
<arg>-ow</arg><arg>xml/works/xml_works_1.seq</arg>
<arg>-oew</arg><arg>---</arg>
</java>
<ok to="join_node"/>
<error to="Kill"/>
</action>
<action name="ExtractXMLWorkActivities_2">
<java>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<main-class>eu.dnetlib.doiboost.orcid.ExtractXMLActivitiesData</main-class>
<arg>-w</arg><arg>${workingPath}/</arg>
<arg>-n</arg><arg>${nameNode}</arg>
<arg>-f</arg><arg>ORCID_2020_10_activites_2.tar.gz</arg>
<arg>-ow</arg><arg>xml/works/xml_works_2.seq</arg>
<arg>-oew</arg><arg>---</arg>
</java>
<ok to="join_node"/>
<error to="Kill"/>
</action>
<action name="ExtractXMLWorkActivities_3">
<java>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<main-class>eu.dnetlib.doiboost.orcid.ExtractXMLActivitiesData</main-class>
<arg>-w</arg><arg>${workingPath}/</arg>
<arg>-n</arg><arg>${nameNode}</arg>
<arg>-f</arg><arg>ORCID_2020_10_activites_3.tar.gz</arg>
<arg>-ow</arg><arg>xml/works/xml_works_3.seq</arg>
<arg>-oew</arg><arg>---</arg>
</java>
<ok to="join_node"/>
<error to="Kill"/>
</action>
<action name="ExtractXMLWorkActivities_4">
<java>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<main-class>eu.dnetlib.doiboost.orcid.ExtractXMLActivitiesData</main-class>
<arg>-w</arg><arg>${workingPath}/</arg>
<arg>-n</arg><arg>${nameNode}</arg>
<arg>-f</arg><arg>ORCID_2020_10_activites_4.tar.gz</arg>
<arg>-ow</arg><arg>xml/works/xml_works_4.seq</arg>
<arg>-oew</arg><arg>---</arg>
</java>
<ok to="join_node"/>
<error to="Kill"/>
</action>
<action name="ExtractXMLWorkActivities_5">
<java>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<main-class>eu.dnetlib.doiboost.orcid.ExtractXMLActivitiesData</main-class>
<arg>-w</arg><arg>${workingPath}/</arg>
<arg>-n</arg><arg>${nameNode}</arg>
<arg>-f</arg><arg>ORCID_2020_10_activites_5.tar.gz</arg>
<arg>-ow</arg><arg>xml/works/xml_works_5.seq</arg>
<arg>-oew</arg><arg>---</arg>
</java>
<ok to="join_node"/>
<error to="Kill"/>
</action>
<action name="ExtractXMLWorkActivities_6">
<java>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<main-class>eu.dnetlib.doiboost.orcid.ExtractXMLActivitiesData</main-class>
<arg>-w</arg><arg>${workingPath}/</arg>
<arg>-n</arg><arg>${nameNode}</arg>
<arg>-f</arg><arg>ORCID_2020_10_activites_6.tar.gz</arg>
<arg>-ow</arg><arg>xml/works/xml_works_6.seq</arg>
<arg>-oew</arg><arg>---</arg>
</java>
<ok to="join_node"/>
<error to="Kill"/>
</action>
<action name="ExtractXMLWorkActivities_7">
<java>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<main-class>eu.dnetlib.doiboost.orcid.ExtractXMLActivitiesData</main-class>
<arg>-w</arg><arg>${workingPath}/</arg>
<arg>-n</arg><arg>${nameNode}</arg>
<arg>-f</arg><arg>ORCID_2020_10_activites_7.tar.gz</arg>
<arg>-ow</arg><arg>xml/works/xml_works_7.seq</arg>
<arg>-oew</arg><arg>---</arg>
</java>
<ok to="join_node"/>
<error to="Kill"/>
</action>
<action name="ExtractXMLWorkActivities_8">
<java>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<main-class>eu.dnetlib.doiboost.orcid.ExtractXMLActivitiesData</main-class>
<arg>-w</arg><arg>${workingPath}/</arg>
<arg>-n</arg><arg>${nameNode}</arg>
<arg>-f</arg><arg>ORCID_2020_10_activites_8.tar.gz</arg>
<arg>-ow</arg><arg>xml/works/xml_works_8.seq</arg>
<arg>-oew</arg><arg>---</arg>
</java>
<ok to="join_node"/>
<error to="Kill"/>
</action>
<action name="ExtractXMLWorkActivities_9">
<java>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<main-class>eu.dnetlib.doiboost.orcid.ExtractXMLActivitiesData</main-class>
<arg>-w</arg><arg>${workingPath}/</arg>
<arg>-n</arg><arg>${nameNode}</arg>
<arg>-f</arg><arg>ORCID_2020_10_activites_9.tar.gz</arg>
<arg>-ow</arg><arg>xml/works/xml_works_9.seq</arg>
<arg>-oew</arg><arg>---</arg>
</java>
<ok to="join_node"/>
<error to="Kill"/>
</action>
<action name="ExtractXMLWorkActivities_X">
<java>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<main-class>eu.dnetlib.doiboost.orcid.ExtractXMLActivitiesData</main-class>
<arg>-w</arg><arg>${workingPath}/</arg>
<arg>-n</arg><arg>${nameNode}</arg>
<arg>-f</arg><arg>ORCID_2020_10_activites_X.tar.gz</arg>
<arg>-ow</arg><arg>xml/works/xml_works_X.seq</arg>
<arg>-oew</arg><arg>---</arg>
</java>
<ok to="join_node"/>
<error to="Kill"/>
</action>
<join name = "join_node" to = "End"/>
<end name="End"/>
</workflow-app>