Generate a separate file to list bootstrap properties

core/src/main/resources/bootstrap/tuningTable.yaml

@@ -0,0 +1,158 @@
+# Copyright (c) 2025, NVIDIA CORPORATION.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+tuningDefinitions:
+  - label: spark.executor.cores
+    description: ''
+    enabled: true
+    level: cluster
+    category: tuning
+  - label: spark.executor.instances
+    description: ''
+    enabled: true
+    level: cluster
+    category: tuning
+  - label: spark.executor.memory
+    description: ''
+    enabled: true
+    level: cluster
+    category: tuning
+  - label: spark.executor.memoryOverhead
+    description: ''
+    enabled: true
+    level: cluster
+    category: tuning
+  - label: spark.executor.memoryOverheadFactor
+    description: ''
+    enabled: true
+    level: cluster
+    category: tuning
+  - label: spark.kubernetes.memoryOverheadFactor
+    description: ''
+    enabled: true
+    level: cluster
+    category: tuning
+  - label: spark.locality.wait
+    description: 'The time to wait to launch a data-local task before giving up and launching it on a less-local node. It is recommended to avoid waiting for a data-local task.'
+    enabled: true
+    level: cluster
+    category: tuning
+    defaultSpark: 3s
+  - label: spark.rapids.filecache.enabled
+    description: ''
+    enabled: true
+    level: job
+  - label: spark.rapids.memory.pinnedPool.size
+    description: ''
+    enabled: true
+    level: cluster
+  - label: spark.rapids.shuffle.multiThreaded.maxBytesInFlight
+    description: ''
+    enabled: true
+    level: cluster
+    category: tuning
+  - label: spark.rapids.shuffle.multiThreaded.reader.threads
+    description: ''
+    enabled: true
+    level: cluster
+    category: tuning
+  - label: spark.rapids.shuffle.multiThreaded.writer.threads
+    description: ''
+    enabled: true
+    level: cluster
+    category: tuning
+  - label: spark.rapids.sql.batchSizeBytes
+    description: ''
+    enabled: true
+    level: job
+    category: tuning
+  - label: spark.rapids.sql.concurrentGpuTasks
+    description: ''
+    enabled: true
+    level: cluster
+    category: tuning
+  - label: spark.rapids.sql.format.parquet.multithreaded.combine.waitTime
+    description: ''
+    enabled: true
+    level: cluster
+    category: tuning
+  - label: spark.rapids.sql.enabled
+    description: 'should be true to enable SQL operations on the GPU.'
+    enabled: true
+    level: cluster
+    category: functionality
+  - label: spark.rapids.sql.multiThreadedRead.numThreads
+    description: ''
+    enabled: true
+    level: cluster
+    category: tuning
+  - label: spark.rapids.sql.reader.multithreaded.combine.sizeBytes
+    description: ''
+    enabled: true
+    level: cluster
+    category: tuning
+  - label: spark.shuffle.manager
+    description: ''
+    enabled: true
+    level: cluster
+  - label: spark.sql.adaptive.enabled
+    description: ''
+    enabled: true
+    level: job
+    category: tuning
+  - label: spark.sql.adaptive.advisoryPartitionSizeInBytes
+    description: ''
+    enabled: true
+    level: job
+    category: tuning
+  - label: spark.sql.adaptive.coalescePartitions.initialPartitionNum
+    description: ''
+    enabled: true
+    level: job
+    category: tuning
+  - label: spark.sql.adaptive.coalescePartitions.minPartitionNum
+    description: ''
+    enabled: true
+    level: job
+    category: tuning
+  - label: spark.sql.adaptive.coalescePartitions.minPartitionSize
+    description: ''
+    enabled: true
+    level: job
+    category: tuning
+  - label: spark.sql.adaptive.coalescePartitions.parallelismFirst
+    description: ''
+    enabled: true
+    level: job
+    category: tuning
+  - label: spark.sql.adaptive.autoBroadcastJoinThreshold
+    description: ''
+    enabled: true
+    level: job
+    category: tuning
+  - label: spark.sql.files.maxPartitionBytes
+    description: ''
+    enabled: true
+    level: job
+    category: tuning
+  - label: spark.sql.shuffle.partitions
+    description: ''
+    enabled: true
+    level: job
+    category: tuning
+  - label: spark.task.resource.gpu.amount
+    description: ''
+    enabled: true
+    level: cluster
+    category: tuning

core/src/main/scala/com/nvidia/spark/rapids/tool/profiling/Profiler.scala

@@ -23,7 +23,7 @@ import scala.collection.mutable.{ArrayBuffer, HashMap}
 import scala.util.control.NonFatal
 
 import com.nvidia.spark.rapids.tool.{AppSummaryInfoBaseProvider, EventLogInfo, EventLogPathProcessor, FailedEventLog, Platform, PlatformFactory, ToolBase}
-import com.nvidia.spark.rapids.tool.tuning.{AutoTuner, ProfilingAutoTunerConfigsProvider}
+import com.nvidia.spark.rapids.tool.tuning.{AutoTuner, ProfilingAutoTunerConfigsProvider, TuningEntryTrait}
 import com.nvidia.spark.rapids.tool.views._
 import org.apache.hadoop.conf.Configuration
 
@@ -416,7 +416,7 @@ class Profiler(hadoopConf: Configuration, appArgs: ProfileArgs, enablePB: Boolea
    */
   private def runAutoTuner(appInfo: Option[ApplicationSummaryInfo],
       driverInfoProvider: DriverLogInfoProvider = BaseDriverLogInfoProvider.noneDriverLog)
-  : (Seq[RecommendedPropertyResult], Seq[RecommendedCommentResult]) = {
+  : (Seq[TuningEntryTrait], Seq[RecommendedCommentResult]) = {
     // only run the auto tuner on GPU event logs for profiling tool right now. There are
     // assumptions made in the code
     if (appInfo.isDefined && appInfo.get.appInfo.head.pluginEnabled) {
@@ -611,10 +611,10 @@ object Profiler {
   val COMBINED_LOG_FILE_NAME_PREFIX = "rapids_4_spark_tools_combined"
   val SUBDIR = "rapids_4_spark_profile"
 
-  def getAutoTunerResultsAsString(props: Seq[RecommendedPropertyResult],
+  def getAutoTunerResultsAsString(props: Seq[TuningEntryTrait],
       comments: Seq[RecommendedCommentResult]): String = {
     val propStr = if (props.nonEmpty) {
-        val propertiesToStr = props.map(_.toString).reduce(_ + "\n" + _)
+        val propertiesToStr = props.map(_.toConfString).reduce(_ + "\n" + _)
         s"\nSpark Properties:\n$propertiesToStr\n"
       } else {
         "Cannot recommend properties. See Comments.\n"