D-Net
/
dnet-hadoop
14
5
Fork 11
Code Issues 14 Pull Requests 13 Projects 2 Releases 2 Wiki Activity

[dedup] increased number of partitions in the consistency phase

This commit is contained in:
Claudio Atzori 2021-11-16 08:43:41 +01:00
parent bafa2990f3
commit 0a727d325d
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-dedup-openaire/src/main/resources/eu/dnetlib/dhp/oa/dedup/consistency/oozie_app/workflow.xml
View File

@ -89,7 +89,7 @@
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners} --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=7680 --conf spark.sql.shuffle.partitions=15000
</spark-opts> </spark-opts>
<arg>--graphBasePath</arg><arg>${graphBasePath}</arg> <arg>--graphBasePath</arg><arg>${graphBasePath}</arg>
<arg>--o</arg><arg>${graphOutputPath}</arg> <arg>--o</arg><arg>${graphOutputPath}</arg>
@ -114,7 +114,7 @@
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners} --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=7680 --conf spark.sql.shuffle.partitions=15000
</spark-opts> </spark-opts>
<arg>--graphInputPath</arg><arg>${graphBasePath}</arg> <arg>--graphInputPath</arg><arg>${graphBasePath}</arg>
<arg>--outputPath</arg><arg>${workingPath}/grouped_entities</arg> <arg>--outputPath</arg><arg>${workingPath}/grouped_entities</arg>