andreas.czerniak
/
BrBETA_dnet-hadoop
forked from D-Net/dnet-hadoop
1
1
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

repartition the join_entities in 24k files

This commit is contained in:
Claudio Atzori 2020-05-27 12:44:01 +02:00
parent 2f1a623d09
commit cfd753217c
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
dhp-workflows/dhp-graph-provision/src/main/resources/eu/dnetlib/dhp/oa/provision/oozie_app/workflow.xml
View File

@ -362,7 +362,7 @@
<arg>--inputGraphRootPath</arg><arg>${inputGraphRootPath}</arg> <arg>--inputGraphRootPath</arg><arg>${inputGraphRootPath}</arg>
<arg>--inputRelatedEntitiesPath</arg><arg>${workingDir}/join_partial</arg> <arg>--inputRelatedEntitiesPath</arg><arg>${workingDir}/join_partial</arg>
<arg>--outputPath</arg><arg>${workingDir}/join_entities</arg> <arg>--outputPath</arg><arg>${workingDir}/join_entities</arg>
<arg>--numPartitions</arg><arg>12000</arg> <arg>--numPartitions</arg><arg>24000</arg>
</spark> </spark>
<ok to="adjancency_lists"/> <ok to="adjancency_lists"/>
<error to="Kill"/> <error to="Kill"/>
@ -386,7 +386,7 @@
--conf spark.sql.shuffle.partitions=7680 --conf spark.sql.shuffle.partitions=7680
--conf spark.network.timeout=${sparkNetworkTimeout} --conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts> </spark-opts>
<arg>--inputPath</arg> <arg>${workingDir}/join_entities</arg> <arg>--inputPath</arg><arg>${workingDir}/join_entities</arg>
<arg>--outputPath</arg><arg>${workingDir}/joined</arg> <arg>--outputPath</arg><arg>${workingDir}/joined</arg>
</spark> </spark>
<ok to="convert_to_xml"/> <ok to="convert_to_xml"/>