sabeel
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Pull Requests Activity

[actionmanager] increased spark.sql.shuffle.partitions for publication, dataset, relation records

This commit is contained in:
Claudio Atzori 2023-03-02 09:11:37 +01:00
parent 6f488547a7
commit db9dad4aa7
3 changed files with 5 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
dhp-workflows/dhp-actionmanager/src/main/resources/eu/dnetlib/dhp/actionmanager/wf/dataset/oozie_app/workflow.xml
View File

@ -107,7 +107,7 @@
--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=2560 --conf spark.sql.shuffle.partitions=7000
</spark-opts> </spark-opts>
<arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/dataset</arg> <arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/dataset</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Dataset</arg> <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Dataset</arg>
@ -159,7 +159,7 @@
--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=2560 --conf spark.sql.shuffle.partitions=7000
</spark-opts> </spark-opts>
<arg>--inputGraphTablePath</arg><arg>${workingDir}/dataset</arg> <arg>--inputGraphTablePath</arg><arg>${workingDir}/dataset</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Dataset</arg> <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Dataset</arg>

4
dhp-workflows/dhp-actionmanager/src/main/resources/eu/dnetlib/dhp/actionmanager/wf/publication/oozie_app/workflow.xml
View File

@ -107,7 +107,7 @@
--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=5000 --conf spark.sql.shuffle.partitions=7000
</spark-opts> </spark-opts>
<arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/publication</arg> <arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/publication</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Publication</arg> <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Publication</arg>
@ -159,7 +159,7 @@
--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=5000 --conf spark.sql.shuffle.partitions=7000
</spark-opts> </spark-opts>
<arg>--inputGraphTablePath</arg><arg>${workingDir}/publication</arg> <arg>--inputGraphTablePath</arg><arg>${workingDir}/publication</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Publication</arg> <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Publication</arg>

2
dhp-workflows/dhp-actionmanager/src/main/resources/eu/dnetlib/dhp/actionmanager/wf/relation/oozie_app/workflow.xml
View File

@ -99,7 +99,7 @@
--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=5000 --conf spark.sql.shuffle.partitions=10000
</spark-opts> </spark-opts>
<arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/relation</arg> <arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/relation</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Relation</arg> <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Relation</arg>