Skip to content

[SPARK-29786][SQL] Fix MetaException when dropping a partition not exists on HDFS #26422

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
Deegue wants to merge 14 commits into from
Closed

[SPARK-29786][SQL] Fix MetaException when dropping a partition not exists on HDFS #26422

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
14 commits
Select commit Hold shift + click to select a range
90233e4
Merge pull request #1 from apache/master
Deegue Oct 22, 2019
7a246d5
Merge pull request #2 from apache/master
Deegue Oct 28, 2019
783211f
Merge pull request #3 from apache/master
Deegue Nov 5, 2019
355496e
Fix drop partition MetaException:File does not exist.
Deegue Nov 7, 2019
74d5984
Add notes.
Deegue Nov 14, 2019
84b1da8
1584531350
Deegue Mar 18, 2020
248b6cf
UT failed #119927, getDataLocation->getPartitionPath
Deegue Mar 20, 2020
c20c390
UT failed #120076, add HIDDEN_FILES_PATH_FILTER
Deegue Mar 20, 2020
9d75854
optimize
Deegue Mar 20, 2020
5a5ff74
UT failed #120087. Hive-0.13 doesn't have `Partition.getPartitionPath…
Deegue Mar 20, 2020
f4b3793
UT failed #120093, purge and ignoreIfNotExists should be false.
Deegue Mar 20, 2020
4d1c35e
isEmptyPath->isExistPath
Deegue Mar 27, 2020
7cf7c6b
remove redundant code
Deegue Mar 27, 2020
f8b96de
restore '!'
Deegue May 29, 2020
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
21 changes: 19 additions & 2 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 @@ -27,8 +27,8 @@ import scala.collection.JavaConverters._
import scala.collection.mutable
import scala.collection.mutable.ArrayBuffer

import org.apache.hadoop.fs.Path
import org.apache.hadoop.hive.common.StatsSetupConst
import org.apache.hadoop.fs.{Path, PathFilter}
import org.apache.hadoop.hive.common.{FileUtils, StatsSetupConst}
import org.apache.hadoop.hive.conf.HiveConf
import org.apache.hadoop.hive.conf.HiveConf.ConfVars
import org.apache.hadoop.hive.metastore.{IMetaStoreClient, TableType => HiveTableType}
Expand Down Expand Up @@ -616,6 +616,12 @@ private[hive] class HiveClientImpl(
shim.createPartitions(client, db, table, parts, ignoreIfExists)
}

@throws[Exception]
private def isExistPath(dirPath: Path): Boolean = {
val inpFs = dirPath.getFileSystem(conf)
inpFs.exists(dirPath)
}

override def dropPartitions(
db: String,
table: String,
Expand All @@ -638,6 +644,17 @@ private[hive] class HiveClientImpl(
s"No partition is dropped. One partition spec '$s' does not exist in table '$table' " +
s"database '$db'")
}
// Check whether the partition we are going to drop is empty.
// We make a dummy one for the empty partition. See [SPARK-29786] for more details.
Copy link
Contributor

@cloud-fan cloud-fan Mar 2, 2020

Choose a reason for hiding this comment

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

is this how hive resolve the problem?

Copy link
Contributor Author

@Deegue Deegue Mar 17, 2020

Choose a reason for hiding this comment

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

is this how hive resolve the problem?

Yes, It's the same method as Hive uses.

Copy link
Contributor

@cloud-fan cloud-fan Mar 17, 2020

Choose a reason for hiding this comment

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

Isn't it bad for performance? i.e. you call fs.exists and fs.listStatus for each partition.

Copy link
Contributor Author

@Deegue Deegue Mar 17, 2020 •
edited
Loading

Choose a reason for hiding this comment

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

Isn't it bad for performance? i.e. you call fs.exists and fs.listStatus for each partition.

Yes, but only affect drop partitions. I think it's necessary and won't take much time to do the check while dropping.

Copy link
Contributor

@cloud-fan cloud-fan Mar 23, 2020

Choose a reason for hiding this comment

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

Can you point to the Hive source code that does the same thing? i.e. create a dummy directory before dropping the partition.

Copy link
Contributor Author

@Deegue Deegue Mar 24, 2020

Choose a reason for hiding this comment

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

Can you point to the Hive source code that does the same thing? i.e. create a dummy directory before dropping the partition.

In Hive 1.x, it's like this.

Copy link
Contributor

@cloud-fan cloud-fan Mar 24, 2020

Choose a reason for hiding this comment

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

is it for DROP PARTITION?

Copy link
Contributor Author

@Deegue Deegue Mar 24, 2020

Choose a reason for hiding this comment

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

is it for DROP PARTITION?

No, it will check every query before executing. Maybe it's better to do the check before all queries?

Copy link
Contributor

@cloud-fan cloud-fan Mar 24, 2020

Choose a reason for hiding this comment

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

Does Spark have a problem to do table scan when partition directory not exist?

Copy link
Contributor Author

@Deegue Deegue Mar 26, 2020 •
edited
Loading

Choose a reason for hiding this comment

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

Does Spark have a problem to do table scan when partition directory not exist?

It's related to #24668, and controlled by spark.sql.files.ignoreMissingFiles.
Spark will check it when listing leaf files.

parts.foreach { partition =>
val partPath = partition.getPath.head
if (!isExistPath(partPath)) {
val fs = partPath.getFileSystem(conf)
fs.mkdirs(partPath)
fs.deleteOnExit(partPath)
}
partition
}
parts.map(_.getValues)
}.distinct
var droppedParts = ArrayBuffer.empty[java.util.List[String]]
Expand Down
23 changes: 22 additions & 1 deletion sql/hive/src/test/scala/org/apache/spark/sql/hive/client/VersionsSuite.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -114,11 +114,12 @@ class VersionsSuite extends SparkFunSuite with Logging {

private var versionSpark: TestHiveVersion = null

private val hadoopConf = new Configuration()

versions.foreach { version =>
test(s"$version: create client") {
client = null
System.gc() // Hack to avoid SEGV on some JVM versions.
val hadoopConf = new Configuration()
hadoopConf.set("test", "success")
// Hive changed the default of datanucleus.schema.autoCreateAll from true to false and
// hive.metastore.schema.verification from false to true since 2.0
Expand Down Expand Up @@ -531,6 +532,26 @@ class VersionsSuite extends SparkFunSuite with Logging {
assert(client.getPartitionOption("default", "src_part", spec).isEmpty)
}

test(s"$version: dropPartitions when file not exists") {
val partitions = (1 to testPartitionCount).map { key2 =>
CatalogTablePartition(Map("key1" -> "1", "key2" -> key2.toString), storageFormat)
}
client.createPartitions("default", "src_part", partitions, ignoreIfExists = true)
val spec = Map("key1" -> "1", "key2" -> "2")
val hiveTable = client.getTable("default", "src_part")
val parts = client.getPartitions(hiveTable, Some(spec))
parts.foreach { partition =>
val partPath = new Path(partition.location)
val fs = partPath.getFileSystem(hadoopConf)
if (fs.exists(partPath)) {
fs.delete(partPath, true)
}
}
client.dropPartitions("default", "src_part", Seq(spec), ignoreIfNotExists = false,
purge = false, retainData = false)
assert(client.getPartitionOption("default", "src_part", spec).isEmpty)
}

///////////////////////////////////////////////////////////////////////////
// Function related API
///////////////////////////////////////////////////////////////////////////
Expand Down