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

[actionmanager] spark specific parameters are grouped into sparkClusterOpts and sparkResourceOpts

This commit is contained in:
Claudio Atzori 2024-08-06 15:04:21 +02:00
parent 8e7ef79ce0
commit 47deb565c1
10 changed files with 204 additions and 374 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

58
dhp-workflows/dhp-actionmanager/src/main/resources/eu/dnetlib/dhp/actionmanager/wf/dataset/oozie_app/workflow.xml
View File

@ -28,39 +28,17 @@
<name>shouldGroupById</name> <name>shouldGroupById</name>
<description>indicates whether the promotion operation should group objects in the graph by id or not</description> <description>indicates whether the promotion operation should group objects in the graph by id or not</description>
</property> </property>
<!-- General oozie workflow properties -->
<property> <property>
<name>sparkDriverMemory</name> <name>sparkClusterOpts</name>
<description>memory for driver process</description> <value>--conf spark.network.timeout=600 --conf spark.extraListeners= --conf spark.sql.queryExecutionListeners= --conf spark.yarn.historyServer.address=http://iis-cdh5-test-m3.ocean.icm.edu.pl:18088 --conf spark.eventLog.dir=hdfs://nameservice1/user/spark/applicationHistory --conf spark.shuffle.service.enabled=false</value>
<description>spark cluster-wide options</description>
</property> </property>
<property> <property>
<name>sparkExecutorMemory</name> <name>sparkResourceOpts</name>
<description>memory for individual executor</description> <value>--executor-memory=6G --conf spark.executor.memoryOverhead=4G --executor-cores=6 --driver-memory=8G --driver-cores=4</value>
</property> <description>spark resource options</description>
<property>
<name>sparkExecutorCores</name>
<description>number of cores used by single executor</description>
</property>
<property>
<name>oozieActionShareLibForSpark2</name>
<description>oozie action sharelib for spark 2.*</description>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
<description>spark 2.* extra listeners classname</description>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
<description>spark 2.* sql query execution listeners classname</description>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<description>spark 2.* yarn history server address</description>
</property>
<property>
<name>spark2EventLogDir</name>
<description>spark 2.* event log dir location</description>
</property> </property>
</parameters> </parameters>
@ -100,14 +78,8 @@
<class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class> <class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class>
<jar>dhp-actionmanager-${projectVersion}.jar</jar> <jar>dhp-actionmanager-${projectVersion}.jar</jar>
<spark-opts> <spark-opts>
--executor-memory=${sparkExecutorMemory} ${sparkClusterOpts}
--executor-cores=${sparkExecutorCores} ${sparkResourceOpts}
--driver-memory=${sparkDriverMemory}
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=7000 --conf spark.sql.shuffle.partitions=7000
</spark-opts> </spark-opts>
<arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/dataset</arg> <arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/dataset</arg>
@ -154,14 +126,8 @@
<class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class> <class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class>
<jar>dhp-actionmanager-${projectVersion}.jar</jar> <jar>dhp-actionmanager-${projectVersion}.jar</jar>
<spark-opts> <spark-opts>
--executor-memory=${sparkExecutorMemory} ${sparkClusterOpts}
--executor-cores=${sparkExecutorCores} ${sparkResourceOpts}
--driver-memory=${sparkDriverMemory}
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=7000 --conf spark.sql.shuffle.partitions=7000
</spark-opts> </spark-opts>
<arg>--inputGraphTablePath</arg><arg>${workingDir}/dataset</arg> <arg>--inputGraphTablePath</arg><arg>${workingDir}/dataset</arg>

48
dhp-workflows/dhp-actionmanager/src/main/resources/eu/dnetlib/dhp/actionmanager/wf/datasource/oozie_app/workflow.xml
View File

@ -20,39 +20,17 @@
<name>mergeAndGetStrategy</name> <name>mergeAndGetStrategy</name>
<description>strategy for merging graph table objects with action payload instances, MERGE_FROM_AND_GET or SELECT_NEWER_AND_GET</description> <description>strategy for merging graph table objects with action payload instances, MERGE_FROM_AND_GET or SELECT_NEWER_AND_GET</description>
</property> </property>
<!-- General oozie workflow properties -->
<property> <property>
<name>sparkDriverMemory</name> <name>sparkClusterOpts</name>
<description>memory for driver process</description> <value>--conf spark.network.timeout=600 --conf spark.extraListeners= --conf spark.sql.queryExecutionListeners= --conf spark.yarn.historyServer.address=http://iis-cdh5-test-m3.ocean.icm.edu.pl:18088 --conf spark.eventLog.dir=hdfs://nameservice1/user/spark/applicationHistory --conf spark.shuffle.service.enabled=false</value>
<description>spark cluster-wide options</description>
</property> </property>
<property> <property>
<name>sparkExecutorMemory</name> <name>sparkResourceOpts</name>
<description>memory for individual executor</description> <value>--executor-memory=6G --conf spark.executor.memoryOverhead=4G --executor-cores=6 --driver-memory=8G --driver-cores=4</value>
</property> <description>spark resource options</description>
<property>
<name>sparkExecutorCores</name>
<description>number of cores used by single executor</description>
</property>
<property>
<name>oozieActionShareLibForSpark2</name>
<description>oozie action sharelib for spark 2.*</description>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
<description>spark 2.* extra listeners classname</description>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
<description>spark 2.* sql query execution listeners classname</description>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<description>spark 2.* yarn history server address</description>
</property>
<property>
<name>spark2EventLogDir</name>
<description>spark 2.* event log dir location</description>
</property> </property>
</parameters> </parameters>
@ -92,14 +70,8 @@
<class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class> <class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class>
<jar>dhp-actionmanager-${projectVersion}.jar</jar> <jar>dhp-actionmanager-${projectVersion}.jar</jar>
<spark-opts> <spark-opts>
--executor-memory=${sparkExecutorMemory} ${sparkClusterOpts}
--executor-cores=${sparkExecutorCores} ${sparkResourceOpts}
--driver-memory=${sparkDriverMemory}
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
<arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/datasource</arg> <arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/datasource</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Datasource</arg> <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Datasource</arg>

44
dhp-workflows/dhp-actionmanager/src/main/resources/eu/dnetlib/dhp/actionmanager/wf/main/oozie_app/workflow.xml
View File

@ -61,39 +61,21 @@
<value>false</value> <value>false</value>
<description>indicates whether the promotion operation should group objects in the graph by id or not</description> <description>indicates whether the promotion operation should group objects in the graph by id or not</description>
</property> </property>
<property>
<name>sparkDriverMemory</name>
<description>memory for driver process</description>
</property>
<property>
<name>sparkExecutorMemory</name>
<description>memory for individual executor</description>
</property>
<property>
<name>sparkExecutorCores</name>
<description>number of cores used by single executor</description>
</property>
<property> <property>
<name>oozieActionShareLibForSpark2</name> <name>oozieActionShareLibForSpark2</name>
<description>oozie action sharelib for spark 2.*</description> <description>oozie action sharelib for spark 2.*</description>
</property> </property>
<!-- General oozie workflow properties -->
<property> <property>
<name>spark2ExtraListeners</name> <name>sparkClusterOpts</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value> <value>--conf spark.network.timeout=600 --conf spark.extraListeners= --conf spark.sql.queryExecutionListeners= --conf spark.yarn.historyServer.address=http://iis-cdh5-test-m3.ocean.icm.edu.pl:18088 --conf spark.eventLog.dir=hdfs://nameservice1/user/spark/applicationHistory --conf spark.shuffle.service.enabled=false</value>
<description>spark 2.* extra listeners classname</description> <description>spark cluster-wide options</description>
</property> </property>
<property> <property>
<name>spark2SqlQueryExecutionListeners</name> <name>sparkResourceOpts</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value> <value>--executor-memory=6G --conf spark.executor.memoryOverhead=4G --executor-cores=6 --driver-memory=8G --driver-cores=4</value>
<description>spark 2.* sql query execution listeners classname</description> <description>spark resource options</description>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<description>spark 2.* yarn history server address</description>
</property>
<property>
<name>spark2EventLogDir</name>
<description>spark 2.* event log dir location</description>
</property> </property>
</parameters> </parameters>
@ -122,14 +104,8 @@
<class>eu.dnetlib.dhp.actionmanager.partition.PartitionActionSetsByPayloadTypeJob</class> <class>eu.dnetlib.dhp.actionmanager.partition.PartitionActionSetsByPayloadTypeJob</class>
<jar>dhp-actionmanager-${projectVersion}.jar</jar> <jar>dhp-actionmanager-${projectVersion}.jar</jar>
<spark-opts> <spark-opts>
--executor-memory=${sparkExecutorMemory} ${sparkClusterOpts}
--executor-cores=${sparkExecutorCores} ${sparkResourceOpts}
--driver-memory=${sparkDriverMemory}
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
<arg>--inputActionSetIds</arg><arg>${inputActionSetIds}</arg> <arg>--inputActionSetIds</arg><arg>${inputActionSetIds}</arg>
<arg>--outputPath</arg><arg>${workingDir}/action_payload_by_type</arg> <arg>--outputPath</arg><arg>${workingDir}/action_payload_by_type</arg>

48
dhp-workflows/dhp-actionmanager/src/main/resources/eu/dnetlib/dhp/actionmanager/wf/organization/oozie_app/workflow.xml
View File

@ -20,39 +20,17 @@
<name>mergeAndGetStrategy</name> <name>mergeAndGetStrategy</name>
<description>strategy for merging graph table objects with action payload instances, MERGE_FROM_AND_GET or SELECT_NEWER_AND_GET</description> <description>strategy for merging graph table objects with action payload instances, MERGE_FROM_AND_GET or SELECT_NEWER_AND_GET</description>
</property> </property>
<!-- General oozie workflow properties -->
<property> <property>
<name>sparkDriverMemory</name> <name>sparkClusterOpts</name>
<description>memory for driver process</description> <value>--conf spark.network.timeout=600 --conf spark.extraListeners= --conf spark.sql.queryExecutionListeners= --conf spark.yarn.historyServer.address=http://iis-cdh5-test-m3.ocean.icm.edu.pl:18088 --conf spark.eventLog.dir=hdfs://nameservice1/user/spark/applicationHistory --conf spark.shuffle.service.enabled=false</value>
<description>spark cluster-wide options</description>
</property> </property>
<property> <property>
<name>sparkExecutorMemory</name> <name>sparkResourceOpts</name>
<description>memory for individual executor</description> <value>--executor-memory=6G --conf spark.executor.memoryOverhead=4G --executor-cores=6 --driver-memory=8G --driver-cores=4</value>
</property> <description>spark resource options</description>
<property>
<name>sparkExecutorCores</name>
<description>number of cores used by single executor</description>
</property>
<property>
<name>oozieActionShareLibForSpark2</name>
<description>oozie action sharelib for spark 2.*</description>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
<description>spark 2.* extra listeners classname</description>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
<description>spark 2.* sql query execution listeners classname</description>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<description>spark 2.* yarn history server address</description>
</property>
<property>
<name>spark2EventLogDir</name>
<description>spark 2.* event log dir location</description>
</property> </property>
</parameters> </parameters>
@ -92,14 +70,8 @@
<class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class> <class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class>
<jar>dhp-actionmanager-${projectVersion}.jar</jar> <jar>dhp-actionmanager-${projectVersion}.jar</jar>
<spark-opts> <spark-opts>
--executor-memory=${sparkExecutorMemory} ${sparkClusterOpts}
--executor-cores=${sparkExecutorCores} ${sparkResourceOpts}
--driver-memory=${sparkDriverMemory}
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
<arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/organization</arg> <arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/organization</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Organization</arg> <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Organization</arg>

59
dhp-workflows/dhp-actionmanager/src/main/resources/eu/dnetlib/dhp/actionmanager/wf/otherresearchproduct/oozie_app/workflow.xml
View File

@ -28,39 +28,17 @@
<name>shouldGroupById</name> <name>shouldGroupById</name>
<description>indicates whether the promotion operation should group objects in the graph by id or not</description> <description>indicates whether the promotion operation should group objects in the graph by id or not</description>
</property> </property>
<!-- General oozie workflow properties -->
<property> <property>
<name>sparkDriverMemory</name> <name>sparkClusterOpts</name>
<description>memory for driver process</description> <value>--conf spark.network.timeout=600 --conf spark.extraListeners= --conf spark.sql.queryExecutionListeners= --conf spark.yarn.historyServer.address=http://iis-cdh5-test-m3.ocean.icm.edu.pl:18088 --conf spark.eventLog.dir=hdfs://nameservice1/user/spark/applicationHistory --conf spark.shuffle.service.enabled=false</value>
<description>spark cluster-wide options</description>
</property> </property>
<property> <property>
<name>sparkExecutorMemory</name> <name>sparkResourceOpts</name>
<description>memory for individual executor</description> <value>--executor-memory=6G --conf spark.executor.memoryOverhead=4G --executor-cores=6 --driver-memory=8G --driver-cores=4</value>
</property> <description>spark resource options</description>
<property>
<name>sparkExecutorCores</name>
<description>number of cores used by single executor</description>
</property>
<property>
<name>oozieActionShareLibForSpark2</name>
<description>oozie action sharelib for spark 2.*</description>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
<description>spark 2.* extra listeners classname</description>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
<description>spark 2.* sql query execution listeners classname</description>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<description>spark 2.* yarn history server address</description>
</property>
<property>
<name>spark2EventLogDir</name>
<description>spark 2.* event log dir location</description>
</property> </property>
</parameters> </parameters>
@ -100,14 +78,9 @@
<class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class> <class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class>
<jar>dhp-actionmanager-${projectVersion}.jar</jar> <jar>dhp-actionmanager-${projectVersion}.jar</jar>
<spark-opts> <spark-opts>
--executor-memory=${sparkExecutorMemory} ${sparkClusterOpts}
--executor-cores=${sparkExecutorCores} ${sparkResourceOpts}
--driver-memory=${sparkDriverMemory} --conf spark.sql.shuffle.partitions=8000
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
<arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/otherresearchproduct</arg> <arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/otherresearchproduct</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.OtherResearchProduct</arg> <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.OtherResearchProduct</arg>
@ -153,14 +126,8 @@
<class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class> <class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class>
<jar>dhp-actionmanager-${projectVersion}.jar</jar> <jar>dhp-actionmanager-${projectVersion}.jar</jar>
<spark-opts> <spark-opts>
--executor-memory=${sparkExecutorMemory} ${sparkClusterOpts}
--executor-cores=${sparkExecutorCores} ${sparkResourceOpts}
--driver-memory=${sparkDriverMemory}
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=8000 --conf spark.sql.shuffle.partitions=8000
</spark-opts> </spark-opts>
<arg>--inputGraphTablePath</arg><arg>${workingDir}/otherresearchproduct</arg> <arg>--inputGraphTablePath</arg><arg>${workingDir}/otherresearchproduct</arg>

101
dhp-workflows/dhp-actionmanager/src/main/resources/eu/dnetlib/dhp/actionmanager/wf/person/oozie_app/workflow.xml Normal file
View File

@ -0,0 +1,101 @@
<workflow-app name="promote_action_payload_for_person_table" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>activePromotePersonActionPayload</name>
<description>when true will promote actions with eu.dnetlib.dhp.schema.oaf.Person payload</description>
</property>
<property>
<name>inputGraphRootPath</name>
<description>root location of input materialized graph</description>
</property>
<property>
<name>inputActionPayloadRootPath</name>
<description>root location of action payloads to promote</description>
</property>
<property>
<name>outputGraphRootPath</name>
<description>root location for output materialized graph</description>
</property>
<property>
<name>mergeAndGetStrategy</name>
<description>strategy for merging graph table objects with action payload instances, MERGE_FROM_AND_GET or SELECT_NEWER_AND_GET</description>
</property>
<!-- General oozie workflow properties -->
<property>
<name>sparkClusterOpts</name>
<value>--conf spark.network.timeout=600 --conf spark.extraListeners= --conf spark.sql.queryExecutionListeners= --conf spark.yarn.historyServer.address=http://iis-cdh5-test-m3.ocean.icm.edu.pl:18088 --conf spark.eventLog.dir=hdfs://nameservice1/user/spark/applicationHistory --conf spark.shuffle.service.enabled=false</value>
<description>spark cluster-wide options</description>
</property>
<property>
<name>sparkResourceOpts</name>
<value>--executor-memory=6G --conf spark.executor.memoryOverhead=4G --executor-cores=6 --driver-memory=8G --driver-cores=4</value>
<description>spark resource options</description>
</property>
</parameters>
<global>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<configuration>
<property>
<name>oozie.action.sharelib.for.spark</name>
<value>${oozieActionShareLibForSpark2}</value>
</property>
</configuration>
</global>
<start to="DecisionPromotePersonActionPayload"/>
<kill name="Kill">
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
</kill>
<decision name="DecisionPromotePersonActionPayload">
<switch>
<case to="PromotePersonActionPayloadForPersonTable">
${(activePromotePersonActionPayload eq "true") and
(fs:exists(concat(concat(concat(concat(wf:conf('nameNode'),'/'),wf:conf('inputActionPayloadRootPath')),'/'),'clazz=eu.dnetlib.dhp.schema.oaf.Person')) eq "true")}
</case>
<default to="SkipPromotePersonActionPayloadForPersonTable"/>
</switch>
</decision>
<action name="PromotePersonActionPayloadForPersonTable">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>PromotePersonActionPayloadForPersonTable</name>
<class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class>
<jar>dhp-actionmanager-${projectVersion}.jar</jar>
<spark-opts>
${sparkClusterOpts}
${sparkResourceOpts}
</spark-opts>
<arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/person</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Person</arg>
<arg>--inputActionPayloadPath</arg><arg>${inputActionPayloadRootPath}/clazz=eu.dnetlib.dhp.schema.oaf.Person</arg>
<arg>--actionPayloadClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Person</arg>
<arg>--outputGraphTablePath</arg><arg>${outputGraphRootPath}/person</arg>
<arg>--mergeAndGetStrategy</arg><arg>${mergeAndGetStrategy}</arg>
<arg>--promoteActionStrategy</arg><arg>${promoteActionStrategy}</arg>
</spark>
<ok to="End"/>
<error to="Kill"/>
</action>
<action name="SkipPromotePersonActionPayloadForPersonTable">
<distcp xmlns="uri:oozie:distcp-action:0.2">
<prepare>
<delete path="${outputGraphRootPath}/person"/>
</prepare>
<arg>-pb</arg>
<arg>${inputGraphRootPath}/person</arg>
<arg>${outputGraphRootPath}/person</arg>
</distcp>
<ok to="End"/>
<error to="Kill"/>
</action>
<end name="End"/>
</workflow-app>

47
dhp-workflows/dhp-actionmanager/src/main/resources/eu/dnetlib/dhp/actionmanager/wf/project/oozie_app/workflow.xml
View File

@ -20,39 +20,16 @@
<name>mergeAndGetStrategy</name> <name>mergeAndGetStrategy</name>
<description>strategy for merging graph table objects with action payload instances, MERGE_FROM_AND_GET or SELECT_NEWER_AND_GET</description> <description>strategy for merging graph table objects with action payload instances, MERGE_FROM_AND_GET or SELECT_NEWER_AND_GET</description>
</property> </property>
<!-- General oozie workflow properties -->
<property> <property>
<name>sparkDriverMemory</name> <name>sparkClusterOpts</name>
<description>memory for driver process</description> <value>--conf spark.network.timeout=600 --conf spark.extraListeners= --conf spark.sql.queryExecutionListeners= --conf spark.yarn.historyServer.address=http://iis-cdh5-test-m3.ocean.icm.edu.pl:18088 --conf spark.eventLog.dir=hdfs://nameservice1/user/spark/applicationHistory --conf spark.shuffle.service.enabled=false</value>
<description>spark cluster-wide options</description>
</property> </property>
<property> <property>
<name>sparkExecutorMemory</name> <name>sparkResourceOpts</name>
<description>memory for individual executor</description> <value>--executor-memory=6G --conf spark.executor.memoryOverhead=4G --executor-cores=6 --driver-memory=8G --driver-cores=4</value>
</property> <description>spark resource options</description>
<property>
<name>sparkExecutorCores</name>
<description>number of cores used by single executor</description>
</property>
<property>
<name>oozieActionShareLibForSpark2</name>
<description>oozie action sharelib for spark 2.*</description>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
<description>spark 2.* extra listeners classname</description>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
<description>spark 2.* sql query execution listeners classname</description>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<description>spark 2.* yarn history server address</description>
</property>
<property>
<name>spark2EventLogDir</name>
<description>spark 2.* event log dir location</description>
</property> </property>
</parameters> </parameters>
@ -92,14 +69,8 @@
<class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class> <class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class>
<jar>dhp-actionmanager-${projectVersion}.jar</jar> <jar>dhp-actionmanager-${projectVersion}.jar</jar>
<spark-opts> <spark-opts>
--executor-memory=${sparkExecutorMemory} ${sparkClusterOpts}
--executor-cores=${sparkExecutorCores} ${sparkResourceOpts}
--driver-memory=${sparkDriverMemory}
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
<arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/project</arg> <arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/project</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Project</arg> <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Project</arg>

62
dhp-workflows/dhp-actionmanager/src/main/resources/eu/dnetlib/dhp/actionmanager/wf/publication/oozie_app/workflow.xml
View File

@ -28,39 +28,17 @@
<name>shouldGroupById</name> <name>shouldGroupById</name>
<description>indicates whether the promotion operation should group objects in the graph by id or not</description> <description>indicates whether the promotion operation should group objects in the graph by id or not</description>
</property> </property>
<!-- General oozie workflow properties -->
<property> <property>
<name>sparkDriverMemory</name> <name>sparkClusterOpts</name>
<description>memory for driver process</description> <value>--conf spark.network.timeout=600 --conf spark.extraListeners= --conf spark.sql.queryExecutionListeners= --conf spark.yarn.historyServer.address=http://iis-cdh5-test-m3.ocean.icm.edu.pl:18088 --conf spark.eventLog.dir=hdfs://nameservice1/user/spark/applicationHistory --conf spark.shuffle.service.enabled=false</value>
<description>spark cluster-wide options</description>
</property> </property>
<property> <property>
<name>sparkExecutorMemory</name> <name>sparkResourceOpts</name>
<description>memory for individual executor</description> <value>--executor-memory=6G --conf spark.executor.memoryOverhead=4G --executor-cores=6 --driver-memory=8G --driver-cores=4</value>
</property> <description>spark resource options</description>
<property>
<name>sparkExecutorCores</name>
<description>number of cores used by single executor</description>
</property>
<property>
<name>oozieActionShareLibForSpark2</name>
<description>oozie action sharelib for spark 2.*</description>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
<description>spark 2.* extra listeners classname</description>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
<description>spark 2.* sql query execution listeners classname</description>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<description>spark 2.* yarn history server address</description>
</property>
<property>
<name>spark2EventLogDir</name>
<description>spark 2.* event log dir location</description>
</property> </property>
</parameters> </parameters>
@ -100,15 +78,9 @@
<class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class> <class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class>
<jar>dhp-actionmanager-${projectVersion}.jar</jar> <jar>dhp-actionmanager-${projectVersion}.jar</jar>
<spark-opts> <spark-opts>
--executor-memory=${sparkExecutorMemory} ${sparkClusterOpts}
--executor-cores=${sparkExecutorCores} ${sparkResourceOpts}
--driver-memory=${sparkDriverMemory} --conf spark.sql.shuffle.partitions=12000
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=15000
</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>
@ -154,15 +126,9 @@
<class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class> <class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class>
<jar>dhp-actionmanager-${projectVersion}.jar</jar> <jar>dhp-actionmanager-${projectVersion}.jar</jar>
<spark-opts> <spark-opts>
--executor-memory=${sparkExecutorMemory} ${sparkClusterOpts}
--executor-cores=${sparkExecutorCores} ${sparkResourceOpts}
--driver-memory=${sparkDriverMemory} --conf spark.sql.shuffle.partitions=12000
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=15000
</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>

50
dhp-workflows/dhp-actionmanager/src/main/resources/eu/dnetlib/dhp/actionmanager/wf/relation/oozie_app/workflow.xml
View File

@ -20,39 +20,17 @@
<name>mergeAndGetStrategy</name> <name>mergeAndGetStrategy</name>
<description>strategy for merging graph table objects with action payload instances, MERGE_FROM_AND_GET or SELECT_NEWER_AND_GET</description> <description>strategy for merging graph table objects with action payload instances, MERGE_FROM_AND_GET or SELECT_NEWER_AND_GET</description>
</property> </property>
<!-- General oozie workflow properties -->
<property> <property>
<name>sparkDriverMemory</name> <name>sparkClusterOpts</name>
<description>memory for driver process</description> <value>--conf spark.network.timeout=600 --conf spark.extraListeners= --conf spark.sql.queryExecutionListeners= --conf spark.yarn.historyServer.address=http://iis-cdh5-test-m3.ocean.icm.edu.pl:18088 --conf spark.eventLog.dir=hdfs://nameservice1/user/spark/applicationHistory --conf spark.shuffle.service.enabled=false</value>
<description>spark cluster-wide options</description>
</property> </property>
<property> <property>
<name>sparkExecutorMemory</name> <name>sparkResourceOpts</name>
<description>memory for individual executor</description> <value>--executor-memory=6G --conf spark.executor.memoryOverhead=4G --executor-cores=6 --driver-memory=8G --driver-cores=4</value>
</property> <description>spark resource options</description>
<property>
<name>sparkExecutorCores</name>
<description>number of cores used by single executor</description>
</property>
<property>
<name>oozieActionShareLibForSpark2</name>
<description>oozie action sharelib for spark 2.*</description>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
<description>spark 2.* extra listeners classname</description>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
<description>spark 2.* sql query execution listeners classname</description>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<description>spark 2.* yarn history server address</description>
</property>
<property>
<name>spark2EventLogDir</name>
<description>spark 2.* event log dir location</description>
</property> </property>
</parameters> </parameters>
@ -92,15 +70,9 @@
<class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class> <class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class>
<jar>dhp-actionmanager-${projectVersion}.jar</jar> <jar>dhp-actionmanager-${projectVersion}.jar</jar>
<spark-opts> <spark-opts>
--executor-memory=${sparkExecutorMemory} ${sparkClusterOpts}
--executor-cores=${sparkExecutorCores} ${sparkResourceOpts}
--driver-memory=${sparkDriverMemory} --conf spark.sql.shuffle.partitions=12000
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=15000
</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>

61
dhp-workflows/dhp-actionmanager/src/main/resources/eu/dnetlib/dhp/actionmanager/wf/software/oozie_app/workflow.xml
View File

@ -28,39 +28,17 @@
<name>shouldGroupById</name> <name>shouldGroupById</name>
<description>indicates whether the promotion operation should group objects in the graph by id or not</description> <description>indicates whether the promotion operation should group objects in the graph by id or not</description>
</property> </property>
<!-- General oozie workflow properties -->
<property> <property>
<name>sparkDriverMemory</name> <name>sparkClusterOpts</name>
<description>memory for driver process</description> <value>--conf spark.network.timeout=600 --conf spark.extraListeners= --conf spark.sql.queryExecutionListeners= --conf spark.yarn.historyServer.address=http://iis-cdh5-test-m3.ocean.icm.edu.pl:18088 --conf spark.eventLog.dir=hdfs://nameservice1/user/spark/applicationHistory --conf spark.shuffle.service.enabled=false</value>
<description>spark cluster-wide options</description>
</property> </property>
<property> <property>
<name>sparkExecutorMemory</name> <name>sparkResourceOpts</name>
<description>memory for individual executor</description> <value>--executor-memory=6G --conf spark.executor.memoryOverhead=4G --executor-cores=6 --driver-memory=8G --driver-cores=4</value>
</property> <description>spark resource options</description>
<property>
<name>sparkExecutorCores</name>
<description>number of cores used by single executor</description>
</property>
<property>
<name>oozieActionShareLibForSpark2</name>
<description>oozie action sharelib for spark 2.*</description>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
<description>spark 2.* extra listeners classname</description>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
<description>spark 2.* sql query execution listeners classname</description>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<description>spark 2.* yarn history server address</description>
</property>
<property>
<name>spark2EventLogDir</name>
<description>spark 2.* event log dir location</description>
</property> </property>
</parameters> </parameters>
@ -100,14 +78,9 @@
<class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class> <class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class>
<jar>dhp-actionmanager-${projectVersion}.jar</jar> <jar>dhp-actionmanager-${projectVersion}.jar</jar>
<spark-opts> <spark-opts>
--executor-memory=${sparkExecutorMemory} ${sparkClusterOpts}
--executor-cores=${sparkExecutorCores} ${sparkResourceOpts}
--driver-memory=${sparkDriverMemory} --conf spark.sql.shuffle.partitions=2000
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
<arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/software</arg> <arg>--inputGraphTablePath</arg><arg>${inputGraphRootPath}/software</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Software</arg> <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Software</arg>
@ -153,15 +126,9 @@
<class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class> <class>eu.dnetlib.dhp.actionmanager.promote.PromoteActionPayloadForGraphTableJob</class>
<jar>dhp-actionmanager-${projectVersion}.jar</jar> <jar>dhp-actionmanager-${projectVersion}.jar</jar>
<spark-opts> <spark-opts>
--executor-memory=${sparkExecutorMemory} ${sparkClusterOpts}
--executor-cores=${sparkExecutorCores} ${sparkResourceOpts}
--driver-memory=${sparkDriverMemory} --conf spark.sql.shuffle.partitions=2000
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=4000
</spark-opts> </spark-opts>
<arg>--inputGraphTablePath</arg><arg>${workingDir}/software</arg> <arg>--inputGraphTablePath</arg><arg>${workingDir}/software</arg>
<arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Software</arg> <arg>--graphTableClassName</arg><arg>eu.dnetlib.dhp.schema.oaf.Software</arg>