Github user gatorsmile commented on a diff in the pull request: https://github.com/apache/spark/pull/14971#discussion_r78305513 --- Diff: sql/hive/src/test/scala/org/apache/spark/sql/hive/StatisticsSuite.scala --- @@ -259,6 +259,156 @@ class StatisticsSuite extends QueryTest with TestHiveSingleton with SQLTestUtils } } + private def createNonPartitionedTable(tabName: String): Option[Statistics] = { + val hiveClient = spark.sharedState.externalCatalog.asInstanceOf[HiveExternalCatalog].client + sql( + s""" + |CREATE TABLE $tabName (key STRING, value STRING) + |STORED AS TEXTFILE + |TBLPROPERTIES ('prop1' = 'val1', 'prop2' = 'val2') + """.stripMargin) + sql(s"INSERT INTO TABLE $tabName SELECT * FROM src") + sql(s"ANALYZE TABLE $tabName COMPUTE STATISTICS") + hiveClient.runSqlHive(s"ANALYZE TABLE $tabName COMPUTE STATISTICS") + val describeResult1 = hiveClient.runSqlHive(s"DESCRIBE FORMATTED $tabName") + + val tableMetadata = + spark.sessionState.catalog.getTableMetadata(TableIdentifier(tabName)).properties + // statistics info is not contained in the metadata of the original table + assert(Seq(StatsSetupConst.COLUMN_STATS_ACCURATE, + StatsSetupConst.NUM_FILES, + StatsSetupConst.NUM_PARTITIONS, + StatsSetupConst.ROW_COUNT, + StatsSetupConst.RAW_DATA_SIZE, + StatsSetupConst.TOTAL_SIZE).forall(!tableMetadata.contains(_))) + + assert(StringUtils.filterPattern(describeResult1, "*numRows\\s+500*").nonEmpty) + checkStats( + tabName, isDataSourceTable = false, hasSizeInBytes = true, expectedRowCounts = Some(500)) + } + + test("alter table rename after analyze table") { + val oldName = "tab1" + val newName = "tab2" + withTable(oldName, newName) { + val fetchedStats1 = createNonPartitionedTable(oldName) + sql(s"ALTER TABLE $oldName RENAME TO $newName") + val fetchedStats2 = checkStats( + newName, isDataSourceTable = false, hasSizeInBytes = true, expectedRowCounts = Some(500)) + assert(fetchedStats1 == fetchedStats2) + + // ALTER TABLE RENAME does not affect the contents of Hive specific statistics + val hiveClient = spark.sharedState.externalCatalog.asInstanceOf[HiveExternalCatalog].client + val describeResult = hiveClient.runSqlHive(s"DESCRIBE FORMATTED $newName") + assert(StringUtils.filterPattern(describeResult, "*numRows\\s+500*").nonEmpty) + } + } + + test("alter table SET TBLPROPERTIES after analyze table") { + val tabName = "tab1" + withTable(tabName) { + val fetchedStats1 = createNonPartitionedTable(tabName) + + sql(s"ALTER TABLE $tabName SET TBLPROPERTIES ('foo' = 'a')") + val fetchedStats2 = checkStats( + tabName, isDataSourceTable = false, hasSizeInBytes = true, expectedRowCounts = Some(500)) + assert(fetchedStats1 == fetchedStats2) + + // ALTER TABLE SET TBLPROPERTIES invalidates some contents of Hive specific statistics + // This is triggered by the Hive alterTable API + val hiveClient = spark.sharedState.externalCatalog.asInstanceOf[HiveExternalCatalog].client + val describeResult = hiveClient.runSqlHive(s"DESCRIBE FORMATTED $tabName") + assert(StringUtils.filterPattern(describeResult, "*numRows\\s+-1*").nonEmpty) + } + } + + test("alter table UNSET TBLPROPERTIES after analyze table") { + val tabName = "tab1" + withTable(tabName) { + val fetchedStats1 = createNonPartitionedTable(tabName) + + sql(s"ALTER TABLE $tabName UNSET TBLPROPERTIES ('prop1')") + val fetchedStats2 = checkStats( + tabName, isDataSourceTable = false, hasSizeInBytes = true, expectedRowCounts = Some(500)) + assert(fetchedStats1 == fetchedStats2) + + // ALTER TABLE UNSET TBLPROPERTIES invalidates some contents of Hive specific statistics + // This is triggered by the Hive alterTable API + val hiveClient = spark.sharedState.externalCatalog.asInstanceOf[HiveExternalCatalog].client + val describeResult = hiveClient.runSqlHive(s"DESCRIBE FORMATTED $tabName") + assert(StringUtils.filterPattern(describeResult, "*numRows\\s+-1*").nonEmpty) + } + } + + test("add/drop partitions - managed table") { --- End diff -- FYI, when we drop partitions of EXTERNAL tables, `ANALYZE TABLE` is unable to exclude them from statistics. This should be fixed with https://issues.apache.org/jira/browse/SPARK-17129, if my understanding is right.
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. --- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org