D-Net
/
dnet-hadoop
14
5
Fork 11
Code Issues 14 Pull Requests 12 Projects 2 Releases 2 Wiki Activity

setting spark.shuffle.sort.bypassMergeThreshold equals to spark.sql.shuffle.partitions, see https://spark.apache.org/docs/2.4.0/configuration.html\#shuffle-behavior

This commit is contained in:
Claudio Atzori 2024-02-13 14:18:25 +01:00
parent 1416f16b35
commit ef7335c867
1 changed files with 16 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
dhp-workflows/dhp-graph-provision/src/main/resources/eu/dnetlib/dhp/oa/provision/oozie_app/workflow.xml
View File

@ -152,6 +152,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=3840
--conf spark.shuffle.sort.bypassMergeThreshold=3840
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${inputGraphRootPath}/relation</arg>
<arg>--outputPath</arg><arg>${workingDir}/relation</arg>
@ -190,6 +191,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=15000
--conf spark.shuffle.sort.bypassMergeThreshold=15000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${workingDir}/relation</arg>
@ -217,6 +219,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=15000
--conf spark.shuffle.sort.bypassMergeThreshold=15000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${workingDir}/relation</arg>
@ -244,6 +247,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=10000
--conf spark.shuffle.sort.bypassMergeThreshold=10000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${workingDir}/relation</arg>
@ -271,6 +275,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=5000
--conf spark.shuffle.sort.bypassMergeThreshold=5000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${workingDir}/relation</arg>
@ -298,6 +303,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=5000
--conf spark.shuffle.sort.bypassMergeThreshold=5000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${workingDir}/relation</arg>
@ -325,6 +331,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=5000
--conf spark.shuffle.sort.bypassMergeThreshold=5000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${workingDir}/relation</arg>
@ -352,6 +359,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=5000
--conf spark.shuffle.sort.bypassMergeThreshold=5000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${workingDir}/relation</arg>
@ -391,6 +399,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=15000
--conf spark.shuffle.sort.bypassMergeThreshold=15000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/publication</arg>
@ -419,6 +428,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=10000
--conf spark.shuffle.sort.bypassMergeThreshold=10000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/dataset</arg>
@ -447,6 +457,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=10000
--conf spark.shuffle.sort.bypassMergeThreshold=10000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/otherresearchproduct</arg>
@ -475,6 +486,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=5000
--conf spark.shuffle.sort.bypassMergeThreshold=5000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/software</arg>
@ -503,6 +515,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=8000
--conf spark.shuffle.sort.bypassMergeThreshold=8000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/datasource</arg>
@ -531,6 +544,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=10000
--conf spark.shuffle.sort.bypassMergeThreshold=10000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/organization</arg>
@ -559,6 +573,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=5000
--conf spark.shuffle.sort.bypassMergeThreshold=5000
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputEntityPath</arg><arg>${inputGraphRootPath}/project</arg>
@ -589,6 +604,7 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=3840
--conf spark.shuffle.sort.bypassMergeThreshold=3840
--conf spark.network.timeout=${sparkNetworkTimeout}
</spark-opts>
<arg>--inputPath</arg><arg>${workingDir}/join_entities</arg>