This is an automated email from the ASF dual-hosted git repository.
chengchengjin pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/incubator-gluten.git
The following commit(s) were added to refs/heads/main by this push:
new 32e5fbae8e [GLUTEN-10713] Eliminate the c2r before iceberg partition
write (#10714)
32e5fbae8e is described below
commit 32e5fbae8ee9e1df6371f6d950495ae3daaf6518
Author: Zouxxyy <[email protected]>
AuthorDate: Fri Sep 19 16:56:52 2025 +0800
[GLUTEN-10713] Eliminate the c2r before iceberg partition write (#10714)
Currently, for iceberg partitioned write, an AQE node is inserted before
the write node, which results in unnecessary c2r and r2c operations being added
when we offload the native write
---
.../gluten/component/CHIcebergComponent.scala | 2 +-
.../gluten/component/VeloxIcebergComponent.scala | 3 +-
.../gluten/execution/IcebergWriteJniWrapper.java | 1 -
.../OffloadIcebergWrite.scala | 3 +-
.../execution/enhanced/VeloxIcebergSuite.scala | 50 ++++++++++++++++++++--
.../gluten/backendsapi/velox/VeloxRuleApi.scala | 3 ++
.../OffloadIcebergScan.scala | 3 +-
.../execution/ColumnarV2TableWriteExec.scala | 3 ++
.../extension/columnar/V2WritePostRule.scala | 40 +++++++++++++++++
.../extension/columnar/validator/Validators.scala | 2 +
10 files changed, 101 insertions(+), 9 deletions(-)
diff --git
a/backends-clickhouse/src-iceberg/main/scala/org/apache/gluten/component/CHIcebergComponent.scala
b/backends-clickhouse/src-iceberg/main/scala/org/apache/gluten/component/CHIcebergComponent.scala
index 8ee694cefd..7da4d1ec90 100644
---
a/backends-clickhouse/src-iceberg/main/scala/org/apache/gluten/component/CHIcebergComponent.scala
+++
b/backends-clickhouse/src-iceberg/main/scala/org/apache/gluten/component/CHIcebergComponent.scala
@@ -18,7 +18,7 @@
package org.apache.gluten.component
import org.apache.gluten.backendsapi.clickhouse.CHBackend
-import org.apache.gluten.execution.OffloadIcebergScan
+import org.apache.gluten.extension.OffloadIcebergScan
import org.apache.gluten.extension.injector.Injector
class CHIcebergComponent extends Component {
diff --git
a/backends-velox/src-iceberg/main/scala/org/apache/gluten/component/VeloxIcebergComponent.scala
b/backends-velox/src-iceberg/main/scala/org/apache/gluten/component/VeloxIcebergComponent.scala
index 775c054ca3..c977e17ab4 100644
---
a/backends-velox/src-iceberg/main/scala/org/apache/gluten/component/VeloxIcebergComponent.scala
+++
b/backends-velox/src-iceberg/main/scala/org/apache/gluten/component/VeloxIcebergComponent.scala
@@ -15,8 +15,9 @@
* limitations under the License.
*/
package org.apache.gluten.component
+
import org.apache.gluten.backendsapi.velox.VeloxBackend
-import org.apache.gluten.execution.{OffloadIcebergScan, OffloadIcebergWrite}
+import org.apache.gluten.extension.{OffloadIcebergScan, OffloadIcebergWrite}
import org.apache.gluten.extension.injector.Injector
class VeloxIcebergComponent extends Component {
diff --git
a/backends-velox/src-iceberg/main/scala/org/apache/gluten/execution/IcebergWriteJniWrapper.java
b/backends-velox/src-iceberg/main/scala/org/apache/gluten/execution/IcebergWriteJniWrapper.java
index 25f2e7697c..a9192952cf 100644
---
a/backends-velox/src-iceberg/main/scala/org/apache/gluten/execution/IcebergWriteJniWrapper.java
+++
b/backends-velox/src-iceberg/main/scala/org/apache/gluten/execution/IcebergWriteJniWrapper.java
@@ -16,7 +16,6 @@
*/
package org.apache.gluten.execution;
-import org.apache.gluten.proto.IcebergPartitionSpec;
import org.apache.gluten.runtime.Runtime;
import org.apache.gluten.runtime.RuntimeAware;
diff --git
a/backends-velox/src-iceberg/main/scala/org/apache/gluten/execution/OffloadIcebergWrite.scala
b/backends-velox/src-iceberg/main/scala/org/apache/gluten/extension/OffloadIcebergWrite.scala
similarity index 94%
rename from
backends-velox/src-iceberg/main/scala/org/apache/gluten/execution/OffloadIcebergWrite.scala
rename to
backends-velox/src-iceberg/main/scala/org/apache/gluten/extension/OffloadIcebergWrite.scala
index 1e2beff366..98135fae49 100644
---
a/backends-velox/src-iceberg/main/scala/org/apache/gluten/execution/OffloadIcebergWrite.scala
+++
b/backends-velox/src-iceberg/main/scala/org/apache/gluten/extension/OffloadIcebergWrite.scala
@@ -14,9 +14,10 @@
* See the License for the specific language governing permissions and
* limitations under the License.
*/
-package org.apache.gluten.execution
+package org.apache.gluten.extension
import org.apache.gluten.config.GlutenConfig
+import org.apache.gluten.execution.{VeloxIcebergAppendDataExec,
VeloxIcebergOverwriteByExpressionExec, VeloxIcebergReplaceDataExec}
import org.apache.gluten.extension.columnar.enumerated.RasOffload
import org.apache.gluten.extension.columnar.heuristic.HeuristicTransform
import org.apache.gluten.extension.columnar.offload.OffloadSingleNode
diff --git
a/backends-velox/src-iceberg/test/scala/org/apache/gluten/execution/enhanced/VeloxIcebergSuite.scala
b/backends-velox/src-iceberg/test/scala/org/apache/gluten/execution/enhanced/VeloxIcebergSuite.scala
index a7dfe4e79e..b7270e69a0 100644
---
a/backends-velox/src-iceberg/test/scala/org/apache/gluten/execution/enhanced/VeloxIcebergSuite.scala
+++
b/backends-velox/src-iceberg/test/scala/org/apache/gluten/execution/enhanced/VeloxIcebergSuite.scala
@@ -16,10 +16,10 @@
*/
package org.apache.gluten.execution.enhanced
-import org.apache.gluten.execution.{IcebergSuite, VeloxIcebergAppendDataExec,
VeloxIcebergOverwriteByExpressionExec, VeloxIcebergReplaceDataExec}
+import org.apache.gluten.execution.{ColumnarToRowExecBase, IcebergSuite,
VeloxIcebergAppendDataExec, VeloxIcebergOverwriteByExpressionExec,
VeloxIcebergReplaceDataExec}
import org.apache.gluten.tags.EnhancedFeaturesTest
-import org.apache.spark.sql.Row
+import org.apache.spark.sql.{DataFrame, Row}
import org.apache.spark.sql.execution.CommandResultExec
import org.apache.spark.sql.gluten.TestUtils
@@ -84,8 +84,8 @@ class VeloxIcebergSuite extends IcebergSuite {
|tblproperties (
| 'format-version' = '2',
| 'write.delete.mode' = 'copy-on-write',
- | 'write.update.mode' = 'copy-on-writ',
- | 'write.merge.mode' = 'copy-on-writ'
+ | 'write.update.mode' = 'copy-on-write',
+ | 'write.merge.mode' = 'copy-on-write'
|);
|""".stripMargin)
@@ -217,4 +217,46 @@ class VeloxIcebergSuite extends IcebergSuite {
)
}
}
+
+ test("check iceberg write c2r") {
+ withTable("iceberg_tbl") {
+ spark.sql("""
+ |create table if not exists iceberg_tbl (a int, pt int)
using iceberg
+ |tblproperties (
+ | 'format-version' = '2',
+ | 'write.delete.mode' = 'copy-on-write',
+ | 'write.update.mode' = 'copy-on-write',
+ | 'write.merge.mode' = 'copy-on-write'
+ |)
+ |partitioned by (pt)
+ |""".stripMargin)
+
+ def checkColumnarToRow(df: DataFrame, num: Int): Unit = {
+ assert(
+ collect(
+
df.queryExecution.executedPlan.asInstanceOf[CommandResultExec].commandPhysicalPlan)
{
+ case p if p.isInstanceOf[ColumnarToRowExecBase] => p
+ }.size == num)
+ }
+
+ // insert partitioned table
+ var df = spark.sql("insert into table iceberg_tbl values (1, 1), (2, 1),
(3, 1), (4, 2)")
+ checkAnswer(
+ spark.sql("select * from iceberg_tbl order by a"),
+ Seq(Row(1, 1), Row(2, 1), Row(3, 1), Row(4, 2)))
+ checkColumnarToRow(df, 0)
+
+ // delete partitioned table
+ df = spark.sql("delete from iceberg_tbl where a = 1")
+ checkAnswer(
+ spark.sql("select * from iceberg_tbl order by a"),
+ Seq(Row(2, 1), Row(3, 1), Row(4, 2)))
+ checkColumnarToRow(df, 0)
+
+ // overwrite partitioned table
+ df = spark.sql("insert overwrite table iceberg_tbl values (5, 1)")
+ checkAnswer(spark.sql("select * from iceberg_tbl order by a"),
Seq(Row(5, 1)))
+ checkColumnarToRow(df, 0)
+ }
+ }
}
diff --git
a/backends-velox/src/main/scala/org/apache/gluten/backendsapi/velox/VeloxRuleApi.scala
b/backends-velox/src/main/scala/org/apache/gluten/backendsapi/velox/VeloxRuleApi.scala
index 50fced37ea..57d86a18d4 100644
---
a/backends-velox/src/main/scala/org/apache/gluten/backendsapi/velox/VeloxRuleApi.scala
+++
b/backends-velox/src/main/scala/org/apache/gluten/backendsapi/velox/VeloxRuleApi.scala
@@ -21,6 +21,7 @@ import org.apache.gluten.config.GlutenConfig
import org.apache.gluten.extension._
import org.apache.gluten.extension.columnar._
import
org.apache.gluten.extension.columnar.MiscColumnarRules.{PreventBatchTypeMismatchInTableCache,
RemoveGlutenTableCacheColumnarToRow, RemoveTopmostColumnarToRow,
RewriteSubqueryBroadcast}
+import org.apache.gluten.extension.columnar.V2WritePostRule
import org.apache.gluten.extension.columnar.enumerated.RasOffload
import org.apache.gluten.extension.columnar.heuristic.{ExpandFallbackPolicy,
HeuristicTransform}
import org.apache.gluten.extension.columnar.offload.{OffloadExchange,
OffloadJoin, OffloadOthers}
@@ -122,6 +123,7 @@ object VeloxRuleApi {
injector.injectPostTransform(c =>
HashAggregateIgnoreNullKeysRule.apply(c.session))
injector.injectPostTransform(_ => CollectLimitTransformerRule())
injector.injectPostTransform(_ => CollectTailTransformerRule())
+ injector.injectPostTransform(_ => V2WritePostRule())
injector.injectPostTransform(c =>
InsertTransitions.create(c.outputsColumnar, VeloxBatchType))
// Gluten columnar: Fallback policies.
@@ -223,6 +225,7 @@ object VeloxRuleApi {
injector.injectPostTransform(c =>
HashAggregateIgnoreNullKeysRule.apply(c.session))
injector.injectPostTransform(_ => CollectLimitTransformerRule())
injector.injectPostTransform(_ => CollectTailTransformerRule())
+ injector.injectPostTransform(_ => V2WritePostRule())
injector.injectPostTransform(c =>
InsertTransitions.create(c.outputsColumnar, VeloxBatchType))
injector.injectPostTransform(c => RemoveTopmostColumnarToRow(c.session,
c.caller.isAqe()))
SparkShimLoader.getSparkShims
diff --git
a/gluten-iceberg/src/main/scala/org/apache/gluten/execution/OffloadIcebergScan.scala
b/gluten-iceberg/src/main/scala/org/apache/gluten/extension/OffloadIcebergScan.scala
similarity index 95%
rename from
gluten-iceberg/src/main/scala/org/apache/gluten/execution/OffloadIcebergScan.scala
rename to
gluten-iceberg/src/main/scala/org/apache/gluten/extension/OffloadIcebergScan.scala
index 44b0597503..c6ff8e8b6c 100644
---
a/gluten-iceberg/src/main/scala/org/apache/gluten/execution/OffloadIcebergScan.scala
+++
b/gluten-iceberg/src/main/scala/org/apache/gluten/extension/OffloadIcebergScan.scala
@@ -14,9 +14,10 @@
* See the License for the specific language governing permissions and
* limitations under the License.
*/
-package org.apache.gluten.execution
+package org.apache.gluten.extension
import org.apache.gluten.config.GlutenConfig
+import org.apache.gluten.execution.IcebergScanTransformer
import org.apache.gluten.extension.columnar.enumerated.RasOffload
import org.apache.gluten.extension.columnar.heuristic.HeuristicTransform
import org.apache.gluten.extension.columnar.offload.OffloadSingleNode
diff --git
a/gluten-substrait/src/main/scala/org/apache/gluten/execution/ColumnarV2TableWriteExec.scala
b/gluten-substrait/src/main/scala/org/apache/gluten/execution/ColumnarV2TableWriteExec.scala
index 7265814de7..0479a2d36d 100644
---
a/gluten-substrait/src/main/scala/org/apache/gluten/execution/ColumnarV2TableWriteExec.scala
+++
b/gluten-substrait/src/main/scala/org/apache/gluten/execution/ColumnarV2TableWriteExec.scala
@@ -26,6 +26,7 @@ import org.apache.spark.rdd.RDD
import org.apache.spark.sql.catalyst.InternalRow
import org.apache.spark.sql.connector.write.{BatchWrite, WriterCommitMessage}
import org.apache.spark.sql.datasources.v2.{DataWritingColumnarBatchSparkTask,
DataWritingColumnarBatchSparkTaskResult, StreamWriterCommitProgressUtil,
WritingColumnarBatchSparkTask}
+import org.apache.spark.sql.execution.SparkPlan
import org.apache.spark.sql.execution.datasources.v2._
import org.apache.spark.sql.execution.metric.SQLMetric
import org.apache.spark.sql.types.StructType
@@ -34,6 +35,8 @@ import org.apache.spark.util.LongAccumulator
trait ColumnarV2TableWriteExec extends V2ExistingTableWriteExec with
ValidatablePlan {
+ def withNewQuery(newQuery: SparkPlan): SparkPlan =
withNewChildInternal(newQuery)
+
protected def createFactory(schema: StructType):
ColumnarBatchDataWriterFactory
override protected def run(): Seq[InternalRow] = {
diff --git
a/gluten-substrait/src/main/scala/org/apache/gluten/extension/columnar/V2WritePostRule.scala
b/gluten-substrait/src/main/scala/org/apache/gluten/extension/columnar/V2WritePostRule.scala
new file mode 100644
index 0000000000..d7a4a4e4cc
--- /dev/null
+++
b/gluten-substrait/src/main/scala/org/apache/gluten/extension/columnar/V2WritePostRule.scala
@@ -0,0 +1,40 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You 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.
+ */
+package org.apache.gluten.extension.columnar
+
+import org.apache.gluten.execution.ColumnarV2TableWriteExec
+
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.execution.SparkPlan
+import org.apache.spark.sql.execution.adaptive.AdaptiveSparkPlanExec
+
+case class V2WritePostRule() extends Rule[SparkPlan] {
+
+ override def apply(plan: SparkPlan): SparkPlan = plan match {
+ case write: ColumnarV2TableWriteExec =>
+ /**
+ * If the columnar write's child is aqe, we make aqe "support columnar",
then aqe itself will
+ * guarantee to generate columnar outputs. thus avoiding the case of
c2r->aqe->r2c->writer.
+ */
+ write.query match {
+ case aqe: AdaptiveSparkPlanExec if !aqe.supportsColumnar =>
+ write.withNewQuery(aqe.copy(supportsColumnar = true))
+ case _ => write
+ }
+ case other => other
+ }
+}
diff --git
a/gluten-substrait/src/main/scala/org/apache/gluten/extension/columnar/validator/Validators.scala
b/gluten-substrait/src/main/scala/org/apache/gluten/extension/columnar/validator/Validators.scala
index 3be033451e..2eda0ee12c 100644
---
a/gluten-substrait/src/main/scala/org/apache/gluten/extension/columnar/validator/Validators.scala
+++
b/gluten-substrait/src/main/scala/org/apache/gluten/extension/columnar/validator/Validators.scala
@@ -158,6 +158,8 @@ object Validators {
case p: ShuffleExchangeExec if !glutenConf.enableColumnarShuffle =>
fail(p)
case p: BroadcastExchangeExec if
!glutenConf.enableColumnarBroadcastExchange => fail(p)
case p: AppendDataExec if !glutenConf.enableAppendData => fail(p)
+ case p: ReplaceDataExec if !glutenConf.enableReplaceData => fail(p)
+ case p: OverwriteByExpressionExec if
!glutenConf.enableOverwriteByExpression => fail(p)
case p @ (_: LocalLimitExec | _: GlobalLimitExec) if
!glutenConf.enableColumnarLimit =>
fail(p)
case p: GenerateExec if !glutenConf.enableColumnarGenerate => fail(p)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
