From 60f48c59b78543a88f0d5be885d05896f0385a9f Mon Sep 17 00:00:00 2001
From: Kuhu Shukla <kuhus@nvidia.com>
Date: Fri, 27 Oct 2023 16:22:00 -0500
Subject: [PATCH]  Profiling tool throws NPE when appInfo is null and unchecked

Signed-off-by: Kuhu Shukla <kuhus@nvidia.com>
---
 .../com/nvidia/spark/rapids/tool/profiling/Analysis.scala | 4 +++-
 .../spark/rapids/tool/profiling/CollectInformation.scala  | 8 ++++----
 2 files changed, 7 insertions(+), 5 deletions(-)

diff --git a/core/src/main/scala/com/nvidia/spark/rapids/tool/profiling/Analysis.scala b/core/src/main/scala/com/nvidia/spark/rapids/tool/profiling/Analysis.scala
index 02243ba37..1dc42b34a 100644
--- a/core/src/main/scala/com/nvidia/spark/rapids/tool/profiling/Analysis.scala
+++ b/core/src/main/scala/com/nvidia/spark/rapids/tool/profiling/Analysis.scala
@@ -390,7 +390,9 @@ class Analysis(apps: Seq[ApplicationInfo]) {
     val allRows = apps.flatMap { app =>
       app.sqlIdToInfo.map { case (sqlId, sqlCase) =>
         SQLDurationExecutorTimeProfileResult(app.index, app.appId, sqlId, sqlCase.duration,
-          sqlCase.hasDatasetOrRDD, app.appInfo.duration, sqlCase.problematic,
+          sqlCase.hasDatasetOrRDD,
+          if (app.appInfo != null) app.appInfo.duration else Option(0L),
+          sqlCase.problematic,
           sqlCase.sqlCpuTimePercent)
       }
     }
diff --git a/core/src/main/scala/com/nvidia/spark/rapids/tool/profiling/CollectInformation.scala b/core/src/main/scala/com/nvidia/spark/rapids/tool/profiling/CollectInformation.scala
index 78a6cd5eb..126fa0bdd 100644
--- a/core/src/main/scala/com/nvidia/spark/rapids/tool/profiling/CollectInformation.scala
+++ b/core/src/main/scala/com/nvidia/spark/rapids/tool/profiling/CollectInformation.scala
@@ -34,13 +34,13 @@ case class StageMetrics(numTasks: Int, duration: String)
 class CollectInformation(apps: Seq[ApplicationInfo]) extends Logging {
 
   def getAppInfo: Seq[AppInfoProfileResults] = {
-    val allRows = apps.map { app =>
+    val allRows = apps.filterNot(_.appInfo == null).map { app =>
       val a = app.appInfo
       AppInfoProfileResults(app.index, a.appName, a.appId,
         a.sparkUser,  a.startTime, a.endTime, a.duration,
         a.durationStr, a.sparkVersion, a.pluginEnabled)
     }
-    if (allRows.size > 0) {
+    if (allRows.nonEmpty) {
       allRows.sortBy(cols => (cols.appIndex))
     } else {
       Seq.empty
@@ -48,11 +48,11 @@ class CollectInformation(apps: Seq[ApplicationInfo]) extends Logging {
   }
 
   def getAppLogPath: Seq[AppLogPathProfileResults] = {
-    val allRows = apps.map { app =>
+    val allRows = apps.filterNot(_.appInfo == null).map { app =>
       val a = app.appInfo
       AppLogPathProfileResults(app.index, a.appName, a.appId, app.eventLogPath)
     }
-    if (allRows.size > 0) {
+    if (allRows.nonEmpty) {
       allRows.sortBy(cols => (cols.appIndex))
     } else {
       Seq.empty