forked from apache/hudi
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
[HUDI-5057] Fix msck repair hudi table (apache#6999)
- Loading branch information
Showing
11 changed files
with
431 additions
and
6 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
170 changes: 170 additions & 0 deletions
170
...rk-common/src/main/scala/org/apache/spark/sql/hudi/command/RepairHoodieTableCommand.scala
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,170 @@ | ||
| /* | ||
| * 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.hudi.command | ||
|
|
||
| import org.apache.hadoop.fs.Path | ||
|
|
||
| import org.apache.hudi.common.table.HoodieTableConfig | ||
|
|
||
| import org.apache.spark.sql.catalyst.TableIdentifier | ||
| import org.apache.spark.sql.catalyst.catalog.CatalogTypes.TablePartitionSpec | ||
| import org.apache.spark.sql.catalyst.catalog._ | ||
| import org.apache.spark.sql.execution.command.PartitionStatistics | ||
| import org.apache.spark.sql.hudi.HoodieSqlCommonUtils | ||
| import org.apache.spark.sql.{AnalysisException, Row, SparkSession} | ||
| import org.apache.spark.util.ThreadUtils | ||
|
|
||
| import java.util.concurrent.TimeUnit.MILLISECONDS | ||
| import scala.util.control.NonFatal | ||
|
|
||
| /** | ||
| * Command for repair hudi table's partitions. | ||
| * Use the methods in HoodieSqlCommonUtils to obtain partitions and stats | ||
| * instead of scanning the file system. | ||
| */ | ||
| case class RepairHoodieTableCommand(tableName: TableIdentifier, | ||
| enableAddPartitions: Boolean, | ||
| enableDropPartitions: Boolean, | ||
| cmd: String = "MSCK REPAIR TABLE") extends HoodieLeafRunnableCommand { | ||
|
|
||
| // These are list of statistics that can be collected quickly without requiring a scan of the data | ||
| // see https://github.com/apache/hive/blob/master/ | ||
| // common/src/java/org/apache/hadoop/hive/common/StatsSetupConst.java | ||
| val NUM_FILES = "numFiles" | ||
| val TOTAL_SIZE = "totalSize" | ||
| val DDL_TIME = "transient_lastDdlTime" | ||
|
|
||
| override def run(spark: SparkSession): Seq[Row] = { | ||
| val catalog = spark.sessionState.catalog | ||
| val table = catalog.getTableMetadata(tableName) | ||
| val tableIdentWithDB = table.identifier.quotedString | ||
| if (table.partitionColumnNames.isEmpty) { | ||
| throw new AnalysisException( | ||
| s"Operation not allowed: $cmd only works on partitioned tables: $tableIdentWithDB") | ||
| } | ||
|
|
||
| if (table.storage.locationUri.isEmpty) { | ||
| throw new AnalysisException(s"Operation not allowed: $cmd only works on table with " + | ||
| s"location provided: $tableIdentWithDB") | ||
| } | ||
|
|
||
| val root = new Path(table.location) | ||
| logInfo(s"Recover all the partitions in $root") | ||
|
|
||
| val hoodieCatalogTable = HoodieCatalogTable(spark, table.identifier) | ||
| val isHiveStyledPartitioning = hoodieCatalogTable.catalogProperties. | ||
| getOrElse(HoodieTableConfig.HIVE_STYLE_PARTITIONING_ENABLE.key, "true").toBoolean | ||
|
|
||
| val partitionSpecsAndLocs: Seq[(TablePartitionSpec, Path)] = hoodieCatalogTable. | ||
| getPartitionPaths.map(partitionPath => { | ||
| var values = partitionPath.split('/') | ||
| if (isHiveStyledPartitioning) { | ||
| values = values.map(_.split('=')(1)) | ||
| } | ||
| (table.partitionColumnNames.zip(values).toMap, new Path(root, partitionPath)) | ||
| }) | ||
|
|
||
| val droppedAmount = if (enableDropPartitions) { | ||
| dropPartitions(catalog, partitionSpecsAndLocs) | ||
| } else 0 | ||
| val addedAmount = if (enableAddPartitions) { | ||
| val total = partitionSpecsAndLocs.length | ||
| val partitionStats = if (spark.sqlContext.conf.gatherFastStats) { | ||
| HoodieSqlCommonUtils.getFilesInPartitions(spark, table, partitionSpecsAndLocs | ||
| .map(_._2.toString)) | ||
| .mapValues(statuses => PartitionStatistics(statuses.length, statuses.map(_.getLen).sum)) | ||
| } else { | ||
| Map.empty[String, PartitionStatistics] | ||
| } | ||
| logInfo(s"Finished to gather the fast stats for all $total partitions.") | ||
| addPartitions(spark, table, partitionSpecsAndLocs, partitionStats) | ||
| total | ||
| } else 0 | ||
| // Updates the table to indicate that its partition metadata is stored in the Hive metastore. | ||
| // This is always the case for Hive format tables, but is not true for Datasource tables created | ||
| // before Spark 2.1 unless they are converted via `msck repair table`. | ||
| spark.sessionState.catalog.alterTable(table.copy(tracksPartitionsInCatalog = true)) | ||
| try { | ||
| spark.catalog.refreshTable(tableIdentWithDB) | ||
| } catch { | ||
| case NonFatal(e) => | ||
| logError(s"Cannot refresh the table '$tableIdentWithDB'. A query of the table " + | ||
| "might return wrong result if the table was cached. To avoid such issue, you should " + | ||
| "uncache the table manually via the UNCACHE TABLE command after table recovering will " + | ||
| "complete fully.", e) | ||
| } | ||
| logInfo(s"Recovered all partitions: added ($addedAmount), dropped ($droppedAmount).") | ||
| Seq.empty[Row] | ||
| } | ||
|
|
||
| private def addPartitions(spark: SparkSession, | ||
| table: CatalogTable, | ||
| partitionSpecsAndLocs: Seq[(TablePartitionSpec, Path)], | ||
| partitionStats: Map[String, PartitionStatistics]): Unit = { | ||
| val total = partitionSpecsAndLocs.length | ||
| var done = 0L | ||
| // Hive metastore may not have enough memory to handle millions of partitions in single RPC, | ||
| // we should split them into smaller batches. Since Hive client is not thread safe, we cannot | ||
| // do this in parallel. | ||
| val batchSize = spark.sparkContext.conf.getInt("spark.sql.addPartitionInBatch.size", 100) | ||
| partitionSpecsAndLocs.iterator.grouped(batchSize).foreach { batch => | ||
| val now = MILLISECONDS.toSeconds(System.currentTimeMillis()) | ||
| val parts = batch.map { case (spec, location) => | ||
| val params = partitionStats.get(location.toString).map { | ||
| case PartitionStatistics(numFiles, totalSize) => | ||
| // This two fast stat could prevent Hive metastore to list the files again. | ||
| Map(NUM_FILES -> numFiles.toString, | ||
| TOTAL_SIZE -> totalSize.toString, | ||
| // Workaround a bug in HiveMetastore that try to mutate a read-only parameters. | ||
| // see metastore/src/java/org/apache/hadoop/hive/metastore/HiveMetaStore.java | ||
| DDL_TIME -> now.toString) | ||
| }.getOrElse(Map.empty) | ||
| // inherit table storage format (possibly except for location) | ||
| CatalogTablePartition( | ||
| spec, | ||
| table.storage.copy(locationUri = Some(location.toUri)), | ||
| params) | ||
| } | ||
| spark.sessionState.catalog.createPartitions(tableName, parts, ignoreIfExists = true) | ||
| done += parts.length | ||
| logDebug(s"Recovered ${parts.length} partitions ($done/$total so far)") | ||
| } | ||
| } | ||
|
|
||
| // Drops the partitions that do not exist in partitionSpecsAndLocs | ||
| private def dropPartitions(catalog: SessionCatalog, | ||
| partitionSpecsAndLocs: Seq[(TablePartitionSpec, Path)]): Int = { | ||
| val dropPartSpecs = ThreadUtils.parmap( | ||
| catalog.listPartitions(tableName), | ||
| "RepairTableCommand: non-existing partitions", | ||
| maxThreads = 8) { partition => | ||
| partition.storage.locationUri.flatMap { uri => | ||
| if (partitionSpecsAndLocs.map(_._2).contains(new Path(uri))) None else Some(partition.spec) | ||
| } | ||
| }.flatten | ||
| catalog.dropPartitions( | ||
| tableName, | ||
| dropPartSpecs, | ||
| ignoreIfNotExists = true, | ||
| purge = false, | ||
| // Since we have already checked that partition directories do not exist, we can avoid | ||
| // additional calls to the file system at the catalog side by setting this flag. | ||
| retainData = true) | ||
| dropPartSpecs.length | ||
| } | ||
| } |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.