[GitHub] asfgit closed pull request #23277: [SPARK-26327][SQL] Bug fix for `FileSourceScanExec` metrics update and name changing

[GitBox <gi...@apache.org>]

asfgit closed pull request #23277: [SPARK-26327][SQL] Bug fix for `FileSourceScanExec` metrics update and name changing
URL: https://github.com/apache/spark/pull/23277
 
 
   

This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:

As this is a foreign pull request (from a fork), the diff is supplied
below (as it won't show otherwise due to GitHub magic):

diff --git a/sql/core/src/main/scala/org/apache/spark/sql/execution/DataSourceScanExec.scala b/sql/core/src/main/scala/org/apache/spark/sql/execution/DataSourceScanExec.scala
index b29d5c76c5f3a..c0fa4e777b49c 100644
--- a/sql/core/src/main/scala/org/apache/spark/sql/execution/DataSourceScanExec.scala
+++ b/sql/core/src/main/scala/org/apache/spark/sql/execution/DataSourceScanExec.scala
@@ -167,19 +167,14 @@ case class FileSourceScanExec(
       partitionSchema = relation.partitionSchema,
       relation.sparkSession.sessionState.conf)
 
+  private var fileListingTime = 0L
+
   @transient private lazy val selectedPartitions: Seq[PartitionDirectory] = {
     val optimizerMetadataTimeNs = relation.location.metadataOpsTimeNs.getOrElse(0L)
     val startTime = System.nanoTime()
     val ret = relation.location.listFiles(partitionFilters, dataFilters)
     val timeTakenMs = ((System.nanoTime() - startTime) + optimizerMetadataTimeNs) / 1000 / 1000
-
-    metrics("numFiles").add(ret.map(_.files.size.toLong).sum)
-    metrics("metadataTime").add(timeTakenMs)
-
-    val executionId = sparkContext.getLocalProperty(SQLExecution.EXECUTION_ID_KEY)
-    SQLMetrics.postDriverMetricUpdates(sparkContext, executionId,
-      metrics("numFiles") :: metrics("metadataTime") :: Nil)
-
+    fileListingTime = timeTakenMs
     ret
   }
 
@@ -291,6 +286,8 @@ case class FileSourceScanExec(
   }
 
   private lazy val inputRDD: RDD[InternalRow] = {
+    // Update metrics for taking effect in both code generation node and normal node.
+    updateDriverMetrics()
     val readFile: (PartitionedFile) => Iterator[InternalRow] =
       relation.fileFormat.buildReaderWithPartitionValues(
         sparkSession = relation.sparkSession,
@@ -316,7 +313,7 @@ case class FileSourceScanExec(
   override lazy val metrics =
     Map("numOutputRows" -> SQLMetrics.createMetric(sparkContext, "number of output rows"),
       "numFiles" -> SQLMetrics.createMetric(sparkContext, "number of files"),
-      "metadataTime" -> SQLMetrics.createMetric(sparkContext, "metadata time (ms)"),
+      "fileListingTime" -> SQLMetrics.createMetric(sparkContext, "file listing time (ms)"),
       "scanTime" -> SQLMetrics.createTimingMetric(sparkContext, "scan time"))
 
   protected override def doExecute(): RDD[InternalRow] = {
@@ -507,6 +504,19 @@ case class FileSourceScanExec(
     }
   }
 
+  /**
+   * Send the updated metrics to driver, while this function calling, selectedPartitions has
+   * been initialized. See SPARK-26327 for more detail.
+   */
+  private def updateDriverMetrics() = {
+    metrics("numFiles").add(selectedPartitions.map(_.files.size.toLong).sum)
+    metrics("fileListingTime").add(fileListingTime)
+
+    val executionId = sparkContext.getLocalProperty(SQLExecution.EXECUTION_ID_KEY)
+    SQLMetrics.postDriverMetricUpdates(sparkContext, executionId,
+      metrics("numFiles") :: metrics("fileListingTime") :: Nil)
+  }
+
   override def doCanonicalize(): FileSourceScanExec = {
     FileSourceScanExec(
       relation,
diff --git a/sql/core/src/test/scala/org/apache/spark/sql/execution/metric/SQLMetricsSuite.scala b/sql/core/src/test/scala/org/apache/spark/sql/execution/metric/SQLMetricsSuite.scala
index 2251607e76af8..4a80638f68858 100644
--- a/sql/core/src/test/scala/org/apache/spark/sql/execution/metric/SQLMetricsSuite.scala
+++ b/sql/core/src/test/scala/org/apache/spark/sql/execution/metric/SQLMetricsSuite.scala
@@ -636,4 +636,19 @@ class SQLMetricsSuite extends SparkFunSuite with SQLMetricsTestUtils with Shared
       assert(filters.head.metrics("numOutputRows").value == 1)
     }
   }
+
+  test("SPARK-26327: FileSourceScanExec metrics") {
+    withTable("testDataForScan") {
+      spark.range(10).selectExpr("id", "id % 3 as p")
+        .write.partitionBy("p").saveAsTable("testDataForScan")
+      // The execution plan only has 1 FileScan node.
+      val df = spark.sql(
+        "SELECT * FROM testDataForScan WHERE p = 1")
+      testSparkPlanMetrics(df, 1, Map(
+        0L -> (("Scan parquet default.testdataforscan", Map(
+          "number of output rows" -> 3L,
+          "number of files" -> 2L))))
+      )
+    }
+  }
 }


 

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services

---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org