Skip to content

[SPARK-42480][SQL] Improve the performance of drop partitions #40069

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.

Sign up for GitHub

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

Closed
wecharyu wants to merge 5 commits into from
Closed

[SPARK-42480][SQL] Improve the performance of drop partitions #40069

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
ac729f6
[SPARK-42480][SQL] Improve the performance of drop partitions
wecharyu Feb 17, 2023
2c7446e
resolve partition name should support special chars
wecharyu Feb 18, 2023
8101e4b
remove hive dependency from catalyst module and fix test
wecharyu Feb 19, 2023
a4acae8
add spark.sql.hive.dropPartitionByName.enabled conf
wecharyu Feb 23, 2023
6a806a6
add rename partition test
wecharyu Feb 24, 2023
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
11 changes: 11 additions & 0 deletions sql/catalyst/src/main/scala/org/apache/spark/sql/util/PartitioningUtils.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,6 +30,7 @@ import org.apache.spark.sql.types.{CharType, DataType, StringType, StructField,
import org.apache.spark.unsafe.types.UTF8String

private[sql] object PartitioningUtils {
private val PATTERN_FOR_KEY_EQ_VAL = "(.+)=(.+)".r

@LuciferYang LuciferYang Feb 17, 2023 •
edited
Loading

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

Too idealistic, not all partition tables follow this rule. For example, we can use
alter table ... partition(...) set location ... to relocate the partition to any directory

@LuciferYang LuciferYang Feb 17, 2023

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

So if the data corresponding to the partition a=1 is stored in dir /1/, will there be a bad case with this pr?

@wecharyu wecharyu Feb 18, 2023

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

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

@LuciferYang Thanks for your review, partition name is always followed this rule in Hive makePartName.
Partition name is only related to partition keys and values, other partition fields like location will not affect it.

@LuciferYang LuciferYang Feb 18, 2023 •
edited
Loading

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

I remember seeing similar cases in the production environment, but I can't remember the details. Need to have tests to check the corner scenes we can think of

cc @wangyum @sunchao FYI


def castPartitionSpec(value: String, dt: DataType, conf: SQLConf): Expression = {
conf.storeAssignmentPolicy match {
Expand Down Expand Up @@ -136,4 +137,14 @@ private[sql] object PartitioningUtils {
partitionColumnNames, tableName)
}
}

/**
* Extract the partition values from a partition name, e.g., if a partition name is
* "region=US/dt=2023-02-18", then we will return an array of values ("US", "2023-02-18").
*/
def partitionNameToValues(name: String): Array[String] = {
name.split("/").map {
case PATTERN_FOR_KEY_EQ_VAL(_, v) => v
}
}
}
8 changes: 4 additions & 4 deletions sql/hive/src/main/scala/org/apache/spark/sql/hive/client/HiveClientImpl.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -65,6 +65,7 @@ import org.apache.spark.sql.hive.HiveExternalCatalog
import org.apache.spark.sql.hive.HiveExternalCatalog.DATASOURCE_SCHEMA
import org.apache.spark.sql.internal.SQLConf
import org.apache.spark.sql.types._
import org.apache.spark.sql.util.PartitioningUtils
import org.apache.spark.util.{CircularBuffer, Utils}

/**
Expand Down Expand Up @@ -679,19 +680,18 @@ private[hive] class HiveClientImpl(
purge: Boolean,
retainData: Boolean): Unit = withHiveState {
// TODO: figure out how to drop multiple partitions in one call
val hiveTable = shim.getTable(client, db, table, true /* throw exception */)
// do the check at first and collect all the matching partitions
val matchingParts =
specs.flatMap { s =>
assert(s.values.forall(_.nonEmpty), s"partition spec '$s' is invalid")
// The provided spec here can be a partial spec, i.e. it will match all partitions
// whose specs are supersets of this partial spec. E.g. If a table has partitions
// (b='1', c='1') and (b='1', c='2'), a partial spec of (b='1') will match both.
val parts = shim.getPartitions(client, hiveTable, s.asJava)
if (parts.isEmpty && !ignoreIfNotExists) {
val partitionNames = shim.getPartitionNames(client, db, table, s.asJava, -1)
if (partitionNames.isEmpty && !ignoreIfNotExists) {
throw new NoSuchPartitionsException(db, table, Seq(s))
}
parts.map(_.getValues)
partitionNames.map(PartitioningUtils.partitionNameToValues(_).toList.asJava)
}.distinct
val droppedParts = ArrayBuffer.empty[java.util.List[String]]
matchingParts.foreach { partition =>
Expand Down