You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@paimon.apache.org by "ulysses-you (via GitHub)" <gi...@apache.org> on 2024/08/21 08:41:09 UTC
[PR] [spark] Support report scan ordering [paimon]
ulysses-you opened a new pull request, #4026:
URL: https://github.com/apache/paimon/pull/4026
<!-- Please specify the module before the PR name: [core] ... or [flink] ... -->
### Purpose
<!-- Linking this pull request to the issue -->
This pr makes `PaimonScan` support `SupportsReportOrdering`. For primary key table, we will do a sort run during write so the primary keys in one file are always ordered.
This can help eliminate local sort for smj/sorted agg etc..
<!-- What is the purpose of the change -->
### Tests
<!-- List UT and IT cases to verify this change -->
add tests
### API and Format
<!-- Does this change affect API or storage format -->
no
### Documentation
<!-- Does this change introduce a new feature -->
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@paimon.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [spark] Support report scan ordering [paimon]
Posted by "JingsongLi (via GitHub)" <gi...@apache.org>.
JingsongLi merged PR #4026:
URL: https://github.com/apache/paimon/pull/4026
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@paimon.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [spark] Support report scan ordering [paimon]
Posted by "ulysses-you (via GitHub)" <gi...@apache.org>.
ulysses-you commented on code in PR #4026:
URL: https://github.com/apache/paimon/pull/4026#discussion_r1730585860
##########
paimon-spark/paimon-spark-common/src/main/scala/org/apache/paimon/spark/PaimonScan.scala:
##########
@@ -72,17 +73,56 @@ case class PaimonScan(
}
}
+ private def shouldDoBucketedScan: Boolean = {
+ !bucketedScanDisabled && conf.v2BucketingEnabled && extractBucketTransform.isDefined
+ }
+
+ // Since Spark 3.3
override def outputPartitioning: Partitioning = {
extractBucketTransform
.map(bucket => new KeyGroupedPartitioning(Array(bucket), lazyInputPartitions.size))
.getOrElse(new UnknownPartitioning(0))
}
- override def getInputPartitions(splits: Array[Split]): Seq[PaimonInputPartition] = {
+ // Since Spark 3.4
+ override def outputOrdering(): Array[SortOrder] = {
if (
- bucketedScanDisabled || !conf.v2BucketingEnabled || extractBucketTransform.isEmpty ||
- splits.exists(!_.isInstanceOf[DataSplit])
+ !shouldDoBucketedScan || lazyInputPartitions.exists(
+ !_.isInstanceOf[PaimonBucketedInputPartition])
) {
+ return Array.empty
+ }
+
+ val primaryKeys = table match {
+ case fileStoreTable: FileStoreTable => fileStoreTable.primaryKeys().asScala
+ case _ => Seq.empty
+ }
+ if (primaryKeys.isEmpty) {
+ return Array.empty
+ }
+
+ val allSplitsKeepOrdering = lazyInputPartitions.toSeq
Review Comment:
thank you @JingsongLi, please correct me if wrong. Per my understanding, if a partition contains multi-splits using merge file read, then all splits should be ordered and never be overlapped ?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@paimon.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [spark] Support report scan ordering [paimon]
Posted by "JingsongLi (via GitHub)" <gi...@apache.org>.
JingsongLi commented on PR #4026:
URL: https://github.com/apache/paimon/pull/4026#issuecomment-2301683324
Thanks @ulysses-you for the contribution.
This optimization looks good to me!
But there is some potential issues.
Perhaps we need to add an interface to the ReadBuilder, because even for primary key tables, the reads may not be ordered. For example, if the deletion vectors mode or read-optimized mode is enabled.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@paimon.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [spark] Support report scan ordering [paimon]
Posted by "ulysses-you (via GitHub)" <gi...@apache.org>.
ulysses-you commented on PR #4026:
URL: https://github.com/apache/paimon/pull/4026#issuecomment-2314359097
thank you @JingsongLi , addressed comments
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@paimon.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [spark] Support report scan ordering [paimon]
Posted by "JingsongLi (via GitHub)" <gi...@apache.org>.
JingsongLi commented on code in PR #4026:
URL: https://github.com/apache/paimon/pull/4026#discussion_r1730572731
##########
paimon-spark/paimon-spark-common/src/main/scala/org/apache/paimon/spark/PaimonScan.scala:
##########
@@ -72,17 +73,56 @@ case class PaimonScan(
}
}
+ private def shouldDoBucketedScan: Boolean = {
+ !bucketedScanDisabled && conf.v2BucketingEnabled && extractBucketTransform.isDefined
+ }
+
+ // Since Spark 3.3
override def outputPartitioning: Partitioning = {
extractBucketTransform
.map(bucket => new KeyGroupedPartitioning(Array(bucket), lazyInputPartitions.size))
.getOrElse(new UnknownPartitioning(0))
}
- override def getInputPartitions(splits: Array[Split]): Seq[PaimonInputPartition] = {
+ // Since Spark 3.4
+ override def outputOrdering(): Array[SortOrder] = {
if (
- bucketedScanDisabled || !conf.v2BucketingEnabled || extractBucketTransform.isEmpty ||
- splits.exists(!_.isInstanceOf[DataSplit])
+ !shouldDoBucketedScan || lazyInputPartitions.exists(
+ !_.isInstanceOf[PaimonBucketedInputPartition])
) {
+ return Array.empty
+ }
+
+ val primaryKeys = table match {
+ case fileStoreTable: FileStoreTable => fileStoreTable.primaryKeys().asScala
+ case _ => Seq.empty
+ }
+ if (primaryKeys.isEmpty) {
+ return Array.empty
+ }
+
+ val allSplitsKeepOrdering = lazyInputPartitions.toSeq
Review Comment:
Maybe we should make sure there is only one split in a bucket?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@paimon.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [spark] Support report scan ordering [paimon]
Posted by "JingsongLi (via GitHub)" <gi...@apache.org>.
JingsongLi commented on code in PR #4026:
URL: https://github.com/apache/paimon/pull/4026#discussion_r1733922848
##########
paimon-core/src/main/java/org/apache/paimon/table/source/DataSplit.java:
##########
@@ -108,6 +108,17 @@ public boolean rawConvertible() {
return rawConvertible;
}
+ /**
+ * To indicate if this `DataSplit` keep the raw table ordering. For example, for the primary key
+ * table, we will do sorted run during write and do merge read during read, so the data is
+ * sorted by the primary keys. Return `false` means the ordering is broken. If a `DataSplit` is
+ * `rawConvertible` then there is no sort merge read, so only if the data file number less than
+ * 2, we can get the correct data ordering.
+ */
+ public boolean keepOrdering() {
+ return !rawConvertible || dataFiles.size() < 2;
Review Comment:
Maybe we don't need to introduce method here, just use `!rawConvertible || dataFiles.size() < 2` in spark.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@paimon.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [spark] Support report scan ordering [paimon]
Posted by "ulysses-you (via GitHub)" <gi...@apache.org>.
ulysses-you closed pull request #4026: [spark] Support report scan ordering
URL: https://github.com/apache/paimon/pull/4026
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@paimon.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [spark] Support report scan ordering [paimon]
Posted by "ulysses-you (via GitHub)" <gi...@apache.org>.
ulysses-you commented on code in PR #4026:
URL: https://github.com/apache/paimon/pull/4026#discussion_r1726585652
##########
paimon-core/src/main/java/org/apache/paimon/table/source/DataSplit.java:
##########
@@ -108,6 +108,17 @@ public boolean rawConvertible() {
return rawConvertible;
}
+ /**
+ * To indicate if this `DataSplit` keep the raw table ordering. For example, for the primary key
+ * table, we will do sorted run during write and do merge read during read, so the data is
+ * sorted by the primary keys. Return `false` means the ordering is broken. If a `DataSplit` is
+ * `rawConvertible` then there is no sort merge read, so only if the data file number less than
+ * 2, we can get the correct data ordering.
+ */
+ public boolean keepOrdering() {
+ return !rawConvertible || dataFiles.size() < 2;
+ }
Review Comment:
cc @JingsongLi , If I get it correctly, this should address your concern ?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@paimon.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [spark] Support report scan ordering [paimon]
Posted by "JingsongLi (via GitHub)" <gi...@apache.org>.
JingsongLi commented on code in PR #4026:
URL: https://github.com/apache/paimon/pull/4026#discussion_r1733922476
##########
paimon-spark/paimon-spark-3.3/src/main/scala/org/apache/paimon/spark/PaimonScan.scala:
##########
@@ -0,0 +1,124 @@
+/*
+ * 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.paimon.spark
+
+import org.apache.paimon.predicate.Predicate
+import org.apache.paimon.table.{BucketMode, FileStoreTable, Table}
+import org.apache.paimon.table.source.{DataSplit, Split}
+
+import org.apache.spark.sql.PaimonUtils.fieldReference
+import org.apache.spark.sql.connector.expressions.{Expressions, NamedReference, Transform}
+import org.apache.spark.sql.connector.read.{SupportsReportPartitioning, SupportsRuntimeFiltering}
+import org.apache.spark.sql.connector.read.partitioning.{KeyGroupedPartitioning, Partitioning, UnknownPartitioning}
+import org.apache.spark.sql.sources.{Filter, In}
+import org.apache.spark.sql.types.StructType
+
+import scala.collection.JavaConverters._
+
+case class PaimonScan(
Review Comment:
Is there possibility to extract a base class `PaimonBucketedScan` to reuse codes?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@paimon.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [spark] Support report scan ordering [paimon]
Posted by "JingsongLi (via GitHub)" <gi...@apache.org>.
JingsongLi commented on code in PR #4026:
URL: https://github.com/apache/paimon/pull/4026#discussion_r1730686410
##########
paimon-spark/paimon-spark-common/src/main/scala/org/apache/paimon/spark/PaimonScan.scala:
##########
@@ -72,17 +73,56 @@ case class PaimonScan(
}
}
+ private def shouldDoBucketedScan: Boolean = {
+ !bucketedScanDisabled && conf.v2BucketingEnabled && extractBucketTransform.isDefined
+ }
+
+ // Since Spark 3.3
override def outputPartitioning: Partitioning = {
extractBucketTransform
.map(bucket => new KeyGroupedPartitioning(Array(bucket), lazyInputPartitions.size))
.getOrElse(new UnknownPartitioning(0))
}
- override def getInputPartitions(splits: Array[Split]): Seq[PaimonInputPartition] = {
+ // Since Spark 3.4
+ override def outputOrdering(): Array[SortOrder] = {
if (
- bucketedScanDisabled || !conf.v2BucketingEnabled || extractBucketTransform.isEmpty ||
- splits.exists(!_.isInstanceOf[DataSplit])
+ !shouldDoBucketedScan || lazyInputPartitions.exists(
+ !_.isInstanceOf[PaimonBucketedInputPartition])
) {
+ return Array.empty
+ }
+
+ val primaryKeys = table match {
+ case fileStoreTable: FileStoreTable => fileStoreTable.primaryKeys().asScala
+ case _ => Seq.empty
+ }
+ if (primaryKeys.isEmpty) {
+ return Array.empty
+ }
+
+ val allSplitsKeepOrdering = lazyInputPartitions.toSeq
Review Comment:
Merging only occurs within a single split.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@paimon.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org