From cd320efa96b5ac73f2ba40d0926c75548dfc9d49 Mon Sep 17 00:00:00 2001
From: Claudio Atzori <claudio.atzori@isti.cnr.it>
Date: Tue, 21 Apr 2020 16:12:20 +0200
Subject: [PATCH] added extra spark options to graph to hive workflow

---
 .../dhp/oa/graph/hive/oozie_app/workflow.xml  | 39 ++++++++++++++++---
 1 file changed, 33 insertions(+), 6 deletions(-)

diff --git a/dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/hive/oozie_app/workflow.xml b/dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/hive/oozie_app/workflow.xml
index 271c7040f..67ca6a64a 100644
--- a/dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/hive/oozie_app/workflow.xml
+++ b/dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/hive/oozie_app/workflow.xml
@@ -21,6 +21,28 @@
             <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>
     </parameters>
 
     <global>
@@ -35,6 +57,10 @@
                 <name>oozie.launcher.mapred.job.queue.name</name>
                 <value>${oozieLauncherQueueName}</value>
             </property>
+            <property>
+                <name>oozie.action.sharelib.for.spark</name>
+                <value>${oozieActionShareLibForSpark2}</value>
+            </property>
         </configuration>
     </global>
 
@@ -52,14 +78,15 @@
             <class>eu.dnetlib.dhp.oa.graph.GraphHiveImporterJob</class>
             <jar>dhp-graph-mapper-${projectVersion}.jar</jar>
             <spark-opts>
-                --executor-memory ${sparkExecutorMemory}
-                --executor-cores ${sparkExecutorCores}
+                --executor-memory=${sparkExecutorMemory}
+                --executor-cores=${sparkExecutorCores}
                 --driver-memory=${sparkDriverMemory}
-                --conf spark.extraListeners="com.cloudera.spark.lineage.NavigatorAppListener"
-                --conf spark.sql.queryExecutionListeners="com.cloudera.spark.lineage.NavigatorQueryListener"
-                --conf spark.sql.warehouse.dir="/user/hive/warehouse"
+                --conf spark.extraListeners=${spark2ExtraListeners}
+                --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
+                --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
+                --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
+                --conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
             </spark-opts>
-            <arg>-mt</arg> <arg>yarn-cluster</arg>
             <arg>--sourcePath</arg><arg>${sourcePath}</arg>
             <arg>--hive_db_name</arg><arg>${hive_db_name}</arg>
             <arg>--hive_metastore_uris</arg><arg>${hive_metastore_uris}</arg>