You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@gluten.apache.org by ma...@apache.org on 2026/07/03 14:51:39 UTC

(gluten) branch main updated: [GLUTEN][CORE] Expose shuffle reader metrics iterator delegate (#12432)

This is an automated email from the ASF dual-hosted git repository.

marin-ma pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/gluten.git


The following commit(s) were added to refs/heads/main by this push:
     new 51a9e96516 [GLUTEN][CORE] Expose shuffle reader metrics iterator delegate (#12432)
51a9e96516 is described below

commit 51a9e96516d4d7414a52027d2019acf46e2b6009
Author: 李扬 <ta...@apache.org>
AuthorDate: Fri Jul 3 22:51:33 2026 +0800

    [GLUTEN][CORE] Expose shuffle reader metrics iterator delegate (#12432)
    
    Change-Id: Ia535725b902e511da1693bd9f7fbaa38f84cc3c8
    
    Co-authored-by: Aime <ai...@bytedance.com>
---
 .../sql/execution/ShuffledColumnarBatchRDD.scala    | 21 ++++++++++++++++-----
 1 file changed, 16 insertions(+), 5 deletions(-)

diff --git a/gluten-substrait/src/main/scala/org/apache/spark/sql/execution/ShuffledColumnarBatchRDD.scala b/gluten-substrait/src/main/scala/org/apache/spark/sql/execution/ShuffledColumnarBatchRDD.scala
index 0642c3a247..fde17aaeed 100644
--- a/gluten-substrait/src/main/scala/org/apache/spark/sql/execution/ShuffledColumnarBatchRDD.scala
+++ b/gluten-substrait/src/main/scala/org/apache/spark/sql/execution/ShuffledColumnarBatchRDD.scala
@@ -28,6 +28,19 @@ import org.apache.spark.sql.vectorized.ColumnarBatch
 final private case class ShuffledColumnarBatchRDDPartition(index: Int, spec: ShufflePartitionSpec)
   extends Partition
 
+/** Wrap metrics with named class, so callers can access the delegate iterator when needed. */
+class ShuffleReaderWithMetricsIterator(
+    val delegate: Iterator[Product2[Int, ColumnarBatch]],
+    sqlMetricsReporter: SQLColumnarShuffleReadMetricsReporter)
+  extends Iterator[ColumnarBatch] {
+  override def hasNext: Boolean = delegate.hasNext
+  override def next(): ColumnarBatch = {
+    val batch = delegate.next()._2
+    sqlMetricsReporter.incBatchesRecordsRead(batch.numRows())
+    batch
+  }
+}
+
 /** [[ShuffledColumnarBatchRDD]] is the columnar version of [[org.apache.spark.rdd.ShuffledRDD]]. */
 class ShuffledColumnarBatchRDD(
     var dependency: ShuffleDependency[Int, ColumnarBatch, ColumnarBatch],
@@ -132,11 +145,9 @@ class ShuffledColumnarBatchRDD(
           context,
           sqlMetricsReporter)
     }
-    reader.read().asInstanceOf[Iterator[Product2[Int, ColumnarBatch]]].map {
-      case (_, batch: ColumnarBatch) =>
-        sqlMetricsReporter.incBatchesRecordsRead(batch.numRows())
-        batch
-    }
+    new ShuffleReaderWithMetricsIterator(
+      reader.read().asInstanceOf[Iterator[Product2[Int, ColumnarBatch]]],
+      sqlMetricsReporter)
   }
 
   override def clearDependencies(): Unit = {


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@gluten.apache.org
For additional commands, e-mail: commits-help@gluten.apache.org