From a2fc37df5f5fe4bccbfc60e99d30658462ed7643 Mon Sep 17 00:00:00 2001
From: Claudio Atzori <claudio.atzori@isti.cnr.it>
Date: Mon, 4 May 2020 19:18:59 +0200
Subject: [PATCH] adjusted parameters

---
 .../oa/graph/raw_db/oozie_app/workflow.xml    | 85 +++++++++++++++----
 1 file changed, 68 insertions(+), 17 deletions(-)

diff --git a/dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/raw_db/oozie_app/workflow.xml b/dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/raw_db/oozie_app/workflow.xml
index 0730f3a1f..05b85a561 100644
--- a/dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/raw_db/oozie_app/workflow.xml
+++ b/dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/raw_db/oozie_app/workflow.xml
@@ -1,8 +1,8 @@
-<workflow-app name="import db entities (step 1)" xmlns="uri:oozie:workflow:0.5">
+<workflow-app name="import DB entities" xmlns="uri:oozie:workflow:0.5">
     <parameters>
         <property>
-            <name>migrationPathStep1</name>
-            <description>the base path to store hdfs file</description>
+            <name>contentPath</name>
+            <description>path location to store (or reuse) content from the aggregator</description>
         </property>
         <property>
             <name>postgresURL</name>
@@ -16,6 +16,7 @@
             <name>postgresPassword</name>
             <description>the password postgres</description>
         </property>
+
         <property>
             <name>sparkDriverMemory</name>
             <description>memory for driver process</description>
@@ -28,31 +29,81 @@
             <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>
 
-	<start to="ResetWorkingPath"/>
+    <global>
+        <job-tracker>${jobTracker}</job-tracker>
+        <name-node>${nameNode}</name-node>
+        <configuration>
+            <property>
+                <name>mapreduce.job.queuename</name>
+                <value>${queueName}</value>
+            </property>
+            <property>
+                <name>oozie.launcher.mapred.job.queue.name</name>
+                <value>${oozieLauncherQueueName}</value>
+            </property>
+            <property>
+                <name>oozie.action.sharelib.for.spark</name>
+                <value>${oozieActionShareLibForSpark2}</value>
+            </property>
+        </configuration>
+    </global>
+
+	<start to="ImportDB"/>
 
     <kill name="Kill">
         <message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
     </kill>
 
-    <action name="ResetWorkingPath">
-        <fs>
-            <delete path='${migrationPathStep1}/db_records'/>
-        </fs>
-        <ok to="ImportDB"/>
+    <action name="ImportDB">
+        <java>
+            <prepare>
+                <delete path="${contentPath}/db_records"/>
+            </prepare>
+            <main-class>eu.dnetlib.dhp.oa.graph.raw.MigrateDbEntitiesApplication</main-class>
+            <arg>--hdfsPath</arg><arg>${contentPath}/db_records</arg>
+            <arg>--postgresUrl</arg><arg>${postgresURL}</arg>
+            <arg>--postgresUser</arg><arg>${postgresUser}</arg>
+            <arg>--postgresPassword</arg><arg>${postgresPassword}</arg>
+        </java>
+        <ok to="ImportDB_claims"/>
         <error to="Kill"/>
     </action>
 
-    <action name="ImportDB">
+    <action name="ImportDB_claims">
         <java>
-            <job-tracker>${jobTracker}</job-tracker>
-            <name-node>${nameNode}</name-node>
-            <main-class>eu.dnetlib.dhp.migration.step1.MigrateDbEntitiesApplication</main-class>
-            <arg>-p</arg><arg>${migrationPathStep1}/db_records</arg>
-            <arg>-pgurl</arg><arg>${postgresURL}</arg>
-            <arg>-pguser</arg><arg>${postgresUser}</arg>
-            <arg>-pgpasswd</arg><arg>${postgresPassword}</arg>
+            <prepare>
+                <delete path="${contentPath}/db_claims"/>
+            </prepare>
+            <main-class>eu.dnetlib.dhp.oa.graph.raw.MigrateDbEntitiesApplication</main-class>
+            <arg>--hdfsPath</arg><arg>${contentPath}/db_claims</arg>
+            <arg>--postgresUrl</arg><arg>${postgresURL}</arg>
+            <arg>--postgresUser</arg><arg>${postgresUser}</arg>
+            <arg>--postgresPassword</arg><arg>${postgresPassword}</arg>
+            <arg>--action</arg><arg>claims</arg>
         </java>
         <ok to="End"/>
         <error to="Kill"/>