przemyslaw.jacewicz
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

added wf nodes to copy entities to the output path

This commit is contained in:
Claudio Atzori 2020-05-11 18:16:39 +02:00
parent c6b028f2af
commit f9a62ba63b
1 changed files with 90 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

91
dhp-workflows/dhp-blacklist/src/main/resources/eu/dnetlib/dhp/blacklist/oozie_app/workflow.xml
View File

@ -34,10 +34,99 @@
<delete path="${outputPath}"/>
<mkdir path="${outputPath}"/>
</fs>
<ok to="read_blacklist"/>
<ok to="copy_entities"/>
<error to="Kill"/>
</action>
<fork name="copy_entities">
<path start="copy_publication"/>
<path start="copy_dataset"/>
<path start="copy_orp"/>
<path start="copy_software"/>
<path start="copy_datasource"/>
<path start="copy_project"/>
<path start="copy_organization"/>
</fork>
<action name="copy_publication">
<distcp xmlns="uri:oozie:distcp-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<arg>${nameNode}/${sourcePath}/publication</arg>
<arg>${nameNode}/${outputPath}/publication</arg>
</distcp>
<ok to="wait"/>
<error to="Kill"/>
</action>
<action name="copy_dataset">
<distcp xmlns="uri:oozie:distcp-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<arg>${nameNode}/${sourcePath}/dataset</arg>
<arg>${nameNode}/${outputPath}/dataset</arg>
</distcp>
<ok to="wait"/>
<error to="Kill"/>
</action>
<action name="copy_orp">
<distcp xmlns="uri:oozie:distcp-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<arg>${nameNode}/${sourcePath}/otherresearchproduct</arg>
<arg>${nameNode}/${outputPath}/otherresearchproduct</arg>
</distcp>
<ok to="wait"/>
<error to="Kill"/>
</action>
<action name="copy_software">
<distcp xmlns="uri:oozie:distcp-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<arg>${nameNode}/${sourcePath}/software</arg>
<arg>${nameNode}/${outputPath}/software</arg>
</distcp>
<ok to="wait"/>
<error to="Kill"/>
</action>
<action name="copy_organization">
<distcp xmlns="uri:oozie:distcp-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<arg>${nameNode}/${sourcePath}/organization</arg>
<arg>${nameNode}/${outputPath}/organization</arg>
</distcp>
<ok to="wait"/>
<error to="Kill"/>
</action>
<action name="copy_projects">
<distcp xmlns="uri:oozie:distcp-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<arg>${nameNode}/${sourcePath}/project</arg>
<arg>${nameNode}/${outputPath}/project</arg>
</distcp>
<ok to="wait"/>
<error to="Kill"/>
</action>
<action name="copy_datasources">
<distcp xmlns="uri:oozie:distcp-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<arg>${nameNode}/${sourcePath}/datasource</arg>
<arg>${nameNode}/${outputPath}/datasource</arg>
</distcp>
<ok to="wait"/>
<error to="Kill"/>
</action>
<join name="wait" to="read_blacklist"/>
<action name="read_blacklist">
<java>
<job-tracker>${jobTracker}</job-tracker>