Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[GLUTEN-3547][CORE] [VL] Add native parquet writer in spark 3.4 #3690

Merged
merged 5 commits into from
Dec 27, 2023
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Original file line number Diff line number Diff line change
Expand Up @@ -30,6 +30,7 @@ import org.apache.spark.sql.catalyst.expressions.aggregate.AggregateExpression
import org.apache.spark.sql.catalyst.plans.physical.{HashPartitioning, Partitioning}
import org.apache.spark.sql.execution.SparkPlan
import org.apache.spark.sql.execution.aggregate.HashAggregateExec
import org.apache.spark.sql.execution.datasources.FileFormat
import org.apache.spark.sql.internal.SQLConf
import org.apache.spark.sql.types.{ArrayType, MapType, StructField, StructType}

Expand Down Expand Up @@ -269,4 +270,8 @@ object CHBackendSettings extends BackendSettingsApi with Logging {
SparkEnv.get.conf
.getLong(GLUTEN_MAX_SHUFFLE_READ_BYTES, GLUTEN_MAX_SHUFFLE_READ_BYTES_DEFAULT)
}

override def supportWriteFilesExec(
format: FileFormat,
fields: Array[StructField]): Option[String] = None
}
Original file line number Diff line number Diff line change
Expand Up @@ -31,6 +31,8 @@ import org.apache.spark.serializer.Serializer
import org.apache.spark.shuffle.{GenShuffleWriterParameters, GlutenShuffleWriterWrapper}
import org.apache.spark.shuffle.utils.CHShuffleUtil
import org.apache.spark.sql.{SparkSession, Strategy}
import org.apache.spark.sql.catalyst.catalog.BucketSpec
import org.apache.spark.sql.catalyst.catalog.CatalogTypes.TablePartitionSpec
import org.apache.spark.sql.catalyst.expressions._
import org.apache.spark.sql.catalyst.expressions.aggregate.AggregateExpression
import org.apache.spark.sql.catalyst.optimizer.BuildSide
Expand All @@ -40,6 +42,7 @@ import org.apache.spark.sql.catalyst.plans.physical.{BroadcastMode, Partitioning
import org.apache.spark.sql.catalyst.rules.Rule
import org.apache.spark.sql.execution._
import org.apache.spark.sql.execution.adaptive.AQEShuffleReadExec
import org.apache.spark.sql.execution.datasources.{FileFormat, WriteFilesExec}
import org.apache.spark.sql.execution.datasources.GlutenWriterColumnarRules.NativeWritePostRule
import org.apache.spark.sql.execution.datasources.v1.ClickHouseFileIndex
import org.apache.spark.sql.execution.datasources.v2.BatchScanExec
Expand Down Expand Up @@ -452,6 +455,16 @@ class CHSparkPlanExecApi extends SparkPlanExecApi {
CHTruncTimestampTransformer(substraitExprName, format, timestamp, timeZoneId, original)
}

override def createColumnarWriteFilesExec(
child: SparkPlan,
fileFormat: FileFormat,
partitionColumns: Seq[Attribute],
bucketSpec: Option[BucketSpec],
options: Map[String, String],
staticPartitions: TablePartitionSpec): WriteFilesExec = {
throw new UnsupportedOperationException("ColumnarWriteFilesExec is not support in ch backend.")
}

/**
* Define whether the join operator is fallback because of the join operator is not supported by
* backend
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -23,6 +23,7 @@ import io.glutenproject.execution._
import io.glutenproject.expression._
import io.glutenproject.expression.ConverterUtils.FunctionConfig
import io.glutenproject.memory.nmm.NativeMemoryManagers
import io.glutenproject.sql.shims.SparkShimLoader
import io.glutenproject.substrait.expression.{ExpressionBuilder, ExpressionNode, IfThenNode}
import io.glutenproject.vectorized.{ColumnarBatchSerializer, ColumnarBatchSerializerJniWrapper}

Expand All @@ -34,14 +35,17 @@ import org.apache.spark.shuffle.utils.ShuffleUtil
import org.apache.spark.sql.{SparkSession, Strategy}
import org.apache.spark.sql.catalyst.{AggregateFunctionRewriteRule, FunctionIdentifier}
import org.apache.spark.sql.catalyst.analysis.FunctionRegistry.FunctionBuilder
import org.apache.spark.sql.catalyst.catalog.BucketSpec
import org.apache.spark.sql.catalyst.catalog.CatalogTypes.TablePartitionSpec
import org.apache.spark.sql.catalyst.expressions.{Attribute, Cast, CreateNamedStruct, ElementAt, Expression, ExpressionInfo, GetArrayItem, GetMapValue, GetStructField, Literal, NamedExpression, StringSplit, StringTrim}
import org.apache.spark.sql.catalyst.expressions.aggregate.{AggregateExpression, HLLAdapter}
import org.apache.spark.sql.catalyst.optimizer.BuildSide
import org.apache.spark.sql.catalyst.plans.JoinType
import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
import org.apache.spark.sql.catalyst.plans.physical.{BroadcastMode, Partitioning}
import org.apache.spark.sql.catalyst.rules.Rule
import org.apache.spark.sql.execution.{ColumnarBuildSideRelation, SparkPlan}
import org.apache.spark.sql.execution.{ColumnarBuildSideRelation, SparkPlan, VeloxColumnarWriteFilesExec}
import org.apache.spark.sql.execution.datasources.{FileFormat, WriteFilesExec}
import org.apache.spark.sql.execution.datasources.GlutenWriterColumnarRules.NativeWritePostRule
import org.apache.spark.sql.execution.exchange.BroadcastExchangeExec
import org.apache.spark.sql.execution.joins.BuildSideRelation
Expand Down Expand Up @@ -245,6 +249,22 @@ class SparkPlanExecApiImpl extends SparkPlanExecApi {
ShuffleUtil.genColumnarShuffleWriter(parameters)
}

override def createColumnarWriteFilesExec(
child: SparkPlan,
fileFormat: FileFormat,
partitionColumns: Seq[Attribute],
bucketSpec: Option[BucketSpec],
options: Map[String, String],
staticPartitions: TablePartitionSpec): WriteFilesExec = {
new VeloxColumnarWriteFilesExec(
child,
fileFormat,
partitionColumns,
bucketSpec,
options,
staticPartitions)
}

/**
* Generate ColumnarBatchSerializer for ColumnarShuffleExchangeExec.
*
Expand Down Expand Up @@ -477,7 +497,12 @@ class SparkPlanExecApiImpl extends SparkPlanExecApi {
* @return
*/
override def genExtendedColumnarPostRules(): List[SparkSession => Rule[SparkPlan]] = {
List(spark => NativeWritePostRule(spark))
val Array(major, minor, _) = SparkShimLoader.getSparkShims.getShimDescriptor.toString.split('.')
if (major.toInt > 3 || (major.toInt == 3 && (minor.toInt >= 4))) {
List()
} else {
List(spark => NativeWritePostRule(spark))
}
Comment on lines +500 to +505
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can we somehow do this in shim? Would that require for lot of effort?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@zhztheplayer Good catch. I will add this in shim in the following PRs. Thanks.

}

/**
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -29,7 +29,8 @@ import org.apache.spark.sql.catalyst.plans.JoinType
import org.apache.spark.sql.execution.{ProjectExec, SparkPlan}
import org.apache.spark.sql.execution.aggregate.HashAggregateExec
import org.apache.spark.sql.execution.command.CreateDataSourceTableAsSelectCommand
import org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelationCommand
import org.apache.spark.sql.execution.datasources.{FileFormat, InsertIntoHadoopFsRelationCommand}
import org.apache.spark.sql.execution.datasources.parquet.ParquetFileFormat
import org.apache.spark.sql.expression.UDFResolver
import org.apache.spark.sql.internal.SQLConf
import org.apache.spark.sql.types._
Expand Down Expand Up @@ -114,6 +115,58 @@ object BackendSettings extends BackendSettingsApi {
}
}

override def supportWriteFilesExec(
format: FileFormat,
fields: Array[StructField]): Option[String] = {
def validateCompressionCodec(): Option[String] = {
// Velox doesn't support brotli and lzo.
val unSupportedCompressions = Set("brotli, lzo")
if (unSupportedCompressions.contains(SQLConf.get.parquetCompressionCodec.toLowerCase())) {
Some("brotli or lzo compression codec is not support in velox backend.")
} else {
None
}
}

// Validate if all types are supported.
def validateDateTypes(fields: Array[StructField]): Option[String] = {
fields.flatMap {
field =>
field.dataType match {
case _: TimestampType => Some("TimestampType")
case struct: StructType if validateDateTypes(struct.fields).nonEmpty =>
Some("StructType(TimestampType)")
case array: ArrayType if array.elementType.isInstanceOf[TimestampType] =>
Some("ArrayType(TimestampType)")
case map: MapType
if map.keyType.isInstanceOf[TimestampType] ||
map.valueType.isInstanceOf[TimestampType] =>
Some("MapType(TimestampType)")
case _ => None
}
}.headOption
}

def validateFieldMetadata(fields: Array[StructField]): Option[String] = {
if (fields.exists(!_.metadata.equals(Metadata.empty))) {
Some("StructField contain the metadata information.")
} else None
}

def validateFileFormat(format: FileFormat): Option[String] = {
format match {
case _: ParquetFileFormat => None
case _: FileFormat => Some("Only parquet fileformat is supported in native write.")
}
}

val fileFormat = validateFileFormat(format)
val metadata = validateFieldMetadata(fields)
val dataTypes = validateDateTypes(fields)
val compression = validateCompressionCodec()
compression.orElse(dataTypes).orElse(metadata).orElse(fileFormat)
}

override def supportExpandExec(): Boolean = true

override def supportSortExec(): Boolean = true
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1,146 @@
/*
* 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.spark.sql.execution

import io.glutenproject.columnarbatch.ColumnarBatches
import io.glutenproject.memory.arrowalloc.ArrowBufferAllocators

import org.apache.spark.internal.io.FileCommitProtocol.TaskCommitMessage
import org.apache.spark.rdd.RDD
import org.apache.spark.sql.catalyst.catalog.BucketSpec
import org.apache.spark.sql.catalyst.catalog.CatalogTypes.TablePartitionSpec
import org.apache.spark.sql.catalyst.expressions.Attribute
import org.apache.spark.sql.connector.write.WriterCommitMessage
import org.apache.spark.sql.execution.datasources.{BasicWriteTaskStats, ExecutedWriteSummary, FileFormat, PartitioningUtils, WriteFilesExec, WriteFilesSpec, WriteTaskResult}
import org.apache.spark.sql.vectorized.ColumnarBatch

import shaded.parquet.com.fasterxml.jackson.databind.ObjectMapper

import scala.collection.mutable

class VeloxColumnarWriteFilesExec(
child: SparkPlan,
fileFormat: FileFormat,
partitionColumns: Seq[Attribute],
bucketSpec: Option[BucketSpec],
options: Map[String, String],
staticPartitions: TablePartitionSpec)
extends WriteFilesExec(
child,
fileFormat,
partitionColumns,
bucketSpec,
options,
staticPartitions) {

override def supportsColumnar(): Boolean = true

override def doExecuteWrite(writeFilesSpec: WriteFilesSpec): RDD[WriterCommitMessage] = {
assert(child.supportsColumnar)

// We need to pass the WritePath to the Velox TableWriter in the doTransform
// method of the WriteTransformer. However, the WritePath is not accessible
// during the planning phase in the WriteTransformer, and can only be obtained
// during the actual execution, specifically in the doExecuteWrite method of
// ColumnarWriteFilesExec, where it is available within the WriteFilesSpec.
// Therefore, we use this hack method to pass the writePath.
child.session.sparkContext.setLocalProperty("writePath", writeFilesSpec.description.path)
child.executeColumnar().mapPartitionsInternal {
iter =>
// Currently, the cb contains three columns: row, fragments, and context.
// The first row in the row column contains the number of written numRows.
// The fragments column contains detailed information about the file writes.
// The json can be as following:
// {
// "inMemoryDataSizeInBytes":0,
// "containsNumberedFileNames":true,
// "onDiskDataSizeInBytes":307,
// "fileWriteInfos":[
// {
// "fileSize":307,
// "writeFileName":
// "Gluten_Stage_1_TID_2_0_2_d1db3b31-4f99-41cb-a4e7-3b8607506168.parquet",
// "targetFileName":
// "Gluten_Stage_1_TID_2_0_2_d1db3b31-4f99-41cb-a4e7-3b8607506168.parquet"
// }
// ],
// "writePath":"file:/home/gluten/spark-warehouse/inserttable/part1=1/part2=1",
// "rowCount":1,
// "targetPath":"file:/home/gluten/spark-warehouse/inserttable/part1=1/part2=1",
// "updateMode":"NEW",
// "name":"part1=1/part2=1"
// }
assert(iter.hasNext)
val cb = iter.next()
val loadedCb = ColumnarBatches.ensureLoaded(ArrowBufferAllocators.contextInstance, cb)
JkSelf marked this conversation as resolved.
Show resolved Hide resolved
val numWrittenRows = loadedCb.column(0).getLong(0)

var updatedPartitions = Set.empty[String]
val addedAbsPathFiles: mutable.Map[String, String] = mutable.Map[String, String]()
var numBytes = 0L
for (i <- 0 until loadedCb.numRows() - 1) {
val fragments = loadedCb.column(1).getUTF8String(i + 1)
val objectMapper = new ObjectMapper()
val jsonObject = objectMapper.readTree(fragments.toString)

val fileWriteInfos = jsonObject.get("fileWriteInfos").elements()
if (jsonObject.get("fileWriteInfos").elements().hasNext) {
val writeInfo = fileWriteInfos.next();
numBytes += writeInfo.get("fileSize").longValue()
// Get partition information.
if (jsonObject.get("name").textValue().nonEmpty) {
val targetFileName = writeInfo.get("targetFileName").textValue()
val partitionDir = jsonObject.get("name").textValue()
updatedPartitions += partitionDir
val tmpOutputPath =
writeFilesSpec.description.path + "/" + partitionDir + "/" + targetFileName
val absOutputPathObject =
writeFilesSpec.description.customPartitionLocations.get(
PartitioningUtils.parsePathFragment(partitionDir))
if (absOutputPathObject.nonEmpty) {
val absOutputPath = absOutputPathObject.get + "/" + targetFileName
addedAbsPathFiles(tmpOutputPath) = absOutputPath
}
}
}
}

// TODO: need to get the partition Internal row?
val stats = BasicWriteTaskStats(Seq.empty, loadedCb.numRows() - 1, numBytes, numWrittenRows)
val summary =
ExecutedWriteSummary(updatedPartitions = updatedPartitions, stats = Seq(stats))

val result = WriteTaskResult(
new TaskCommitMessage(addedAbsPathFiles.toMap -> updatedPartitions),
summary)
Iterator.single(result)
}
}

override protected def doExecuteColumnar(): RDD[ColumnarBatch] = {
throw new UnsupportedOperationException(s"This operator doesn't support doExecuteColumnar().")
}

override protected def withNewChildInternal(newChild: SparkPlan): WriteFilesExec =
new VeloxColumnarWriteFilesExec(
newChild,
fileFormat,
partitionColumns,
bucketSpec,
options,
staticPartitions)
}
Original file line number Diff line number Diff line change
Expand Up @@ -445,16 +445,20 @@ class VeloxParquetDataTypeValidationSuite extends VeloxWholeStageTransformerSuit
}
}

ignore("Velox Parquet Write") {
test("Velox Parquet Write") {
withSQLConf(("spark.gluten.sql.native.writer.enabled", "true")) {
withTempDir {
dir =>
val write_path = dir.toURI.getPath
val data_path = getClass.getResource("/").getPath + "/data-type-validation-data/type1"
val df = spark.read.format("parquet").load(data_path)
// Velox native write doesn't support Timestamp type.
val df = spark.read.format("parquet").load(data_path).drop("timestamp")
JkSelf marked this conversation as resolved.
Show resolved Hide resolved
df.write.mode("append").format("parquet").save(write_path)
val parquetDf = spark.read
.format("parquet")
.load(write_path)
checkAnswer(parquetDf, df)
}
}

}
}
Loading
Loading