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

added new parameter in the doiboost process workflow to specify a folder for the process of MAG dataset

This commit is contained in:
Miriam Baglioni 2021-10-13 11:34:24 +02:00 committed by Claudio Atzori
parent 2f61054cd1
commit 8db39c86e2
1 changed files with 6 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
dhp-workflows/dhp-doiboost/src/main/resources/eu/dnetlib/dhp/doiboost/process/oozie_app/workflow.xml
View File

@ -48,7 +48,11 @@
<!-- MAG Parameters --> <!-- MAG Parameters -->
<property> <property>
<name>inputPathMAG</name> <name>inputPathMAG</name>
<description>the MAG working path</description> <description>the MAG input path</description>
</property>
<property>
<name>workingPathMAG</name>
<description>the MAG working path where to store the intermediate process results</description>
</property> </property>
@ -138,7 +142,7 @@
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
<arg>--sourcePath</arg><arg>${inputPathMAG}/dataset</arg> <arg>--sourcePath</arg><arg>${inputPathMAG}/dataset</arg>
<arg>--workingPath</arg><arg>${inputPathMAG}/process_p</arg> <arg>--workingPath</arg><arg>${workingPathMAG}</arg>
<arg>--targetPath</arg><arg>${workingPath}</arg> <arg>--targetPath</arg><arg>${workingPath}</arg>
<arg>--master</arg><arg>yarn-cluster</arg> <arg>--master</arg><arg>yarn-cluster</arg>
</spark> </spark>