Github user gatorsmile commented on a diff in the pull request: https://github.com/apache/spark/pull/18421#discussion_r128296684 --- Diff: sql/hive/src/test/scala/org/apache/spark/sql/hive/StatisticsSuite.scala --- @@ -182,6 +183,189 @@ class StatisticsSuite extends StatisticsCollectionTestBase with TestHiveSingleto } } + test("analyze single partition") { + val tableName = "analyzeTable_part" + + def queryStats(ds: String): CatalogStatistics = { + val partition = + spark.sessionState.catalog.getPartition(TableIdentifier(tableName), Map("ds" -> ds)) + partition.stats.get + } + + def createPartition(ds: String, query: String): Unit = { + sql(s"INSERT INTO TABLE $tableName PARTITION (ds='$ds') $query") + } + + withTable(tableName) { + sql(s"CREATE TABLE $tableName (key STRING, value STRING) PARTITIONED BY (ds STRING)") + + createPartition("2010-01-01", "SELECT '1', 'A' from src") + createPartition("2010-01-02", "SELECT '1', 'A' from src UNION ALL SELECT '1', 'A' from src") + createPartition("2010-01-03", "SELECT '1', 'A' from src") + + sql(s"ANALYZE TABLE $tableName PARTITION (ds='2010-01-01') COMPUTE STATISTICS NOSCAN") + + sql(s"ANALYZE TABLE $tableName PARTITION (ds='2010-01-02') COMPUTE STATISTICS NOSCAN") + + assert(queryStats("2010-01-01").rowCount === None) + assert(queryStats("2010-01-01").sizeInBytes === 2000) + + assert(queryStats("2010-01-02").rowCount === None) + assert(queryStats("2010-01-02").sizeInBytes === 2*2000) + + sql(s"ANALYZE TABLE $tableName PARTITION (ds='2010-01-01') COMPUTE STATISTICS") + + sql(s"ANALYZE TABLE $tableName PARTITION (ds='2010-01-02') COMPUTE STATISTICS") + + assert(queryStats("2010-01-01").rowCount.get === 500) + assert(queryStats("2010-01-01").sizeInBytes === 2000) + + assert(queryStats("2010-01-02").rowCount.get === 2*500) + assert(queryStats("2010-01-02").sizeInBytes === 2*2000) + } + } + + test("analyze a set of partitions") { + val tableName = "analyzeTable_part" + + def queryStats(ds: String, hr: String): Option[CatalogStatistics] = { + val tableId = TableIdentifier(tableName) + val partition = + spark.sessionState.catalog.getPartition(tableId, Map("ds" -> ds, "hr" -> hr)) + partition.stats + } + + def assertPartitionStats( + ds: String, + hr: String, + rowCount: Option[BigInt], + sizeInBytes: BigInt): Unit = { + val stats = queryStats(ds, hr).get + assert(stats.rowCount === rowCount) + assert(stats.sizeInBytes === sizeInBytes) + } + + def createPartition(ds: String, hr: Int, query: String): Unit = { + sql(s"INSERT INTO TABLE $tableName PARTITION (ds='$ds', hr=$hr) $query") + } + + withTable(tableName) { + sql(s"CREATE TABLE $tableName (key STRING, value STRING) PARTITIONED BY (ds STRING, hr INT)") + + createPartition("2010-01-01", 10, "SELECT '1', 'A' from src") + createPartition("2010-01-01", 11, "SELECT '1', 'A' from src") + createPartition("2010-01-02", 10, "SELECT '1', 'A' from src") + createPartition("2010-01-02", 11, + "SELECT '1', 'A' from src UNION ALL SELECT '1', 'A' from src") + + sql(s"ANALYZE TABLE $tableName PARTITION (ds='2010-01-01') COMPUTE STATISTICS NOSCAN") + + assertPartitionStats("2010-01-01", "10", rowCount = None, sizeInBytes = 2000) + assertPartitionStats("2010-01-01", "11", rowCount = None, sizeInBytes = 2000) + assert(queryStats("2010-01-02", "10") === None) + assert(queryStats("2010-01-02", "11") === None) + + sql(s"ANALYZE TABLE $tableName PARTITION (ds='2010-01-02') COMPUTE STATISTICS NOSCAN") + + assertPartitionStats("2010-01-01", "10", rowCount = None, sizeInBytes = 2000) + assertPartitionStats("2010-01-01", "11", rowCount = None, sizeInBytes = 2000) + assertPartitionStats("2010-01-02", "10", rowCount = None, sizeInBytes = 2000) + assertPartitionStats("2010-01-02", "11", rowCount = None, sizeInBytes = 2*2000) + + sql(s"ANALYZE TABLE $tableName PARTITION (ds='2010-01-01') COMPUTE STATISTICS") + + assertPartitionStats("2010-01-01", "10", rowCount = Some(500), sizeInBytes = 2000) + assertPartitionStats("2010-01-01", "11", rowCount = Some(500), sizeInBytes = 2000) + assertPartitionStats("2010-01-02", "10", rowCount = None, sizeInBytes = 2000) + assertPartitionStats("2010-01-02", "11", rowCount = None, sizeInBytes = 2*2000) + + sql(s"ANALYZE TABLE $tableName PARTITION (ds='2010-01-02') COMPUTE STATISTICS") + + assertPartitionStats("2010-01-01", "10", rowCount = Some(500), sizeInBytes = 2000) + assertPartitionStats("2010-01-01", "11", rowCount = Some(500), sizeInBytes = 2000) + assertPartitionStats("2010-01-02", "10", rowCount = Some(500), sizeInBytes = 2000) + assertPartitionStats("2010-01-02", "11", rowCount = Some(2*500), sizeInBytes = 2*2000) + } + } + + test("analyze all partitions") { + val tableName = "analyzeTable_part" + + def assertPartitionStats( + ds: String, + hr: String, + rowCount: Option[BigInt], + sizeInBytes: BigInt): Unit = { + val stats = spark.sessionState.catalog.getPartition(TableIdentifier(tableName), + Map("ds" -> ds, "hr" -> hr)).stats.get + assert(stats.rowCount === rowCount) + assert(stats.sizeInBytes === sizeInBytes) + } + + def createPartition(ds: String, hr: Int, query: String): Unit = { + sql(s"INSERT INTO TABLE $tableName PARTITION (ds='$ds', hr=$hr) $query") + } + + withTable(tableName) { + sql(s"CREATE TABLE $tableName (key STRING, value STRING) PARTITIONED BY (ds STRING, hr INT)") + + createPartition("2010-01-01", 10, "SELECT '1', 'A' from src") + createPartition("2010-01-01", 11, "SELECT '1', 'A' from src") + createPartition("2010-01-02", 10, "SELECT '1', 'A' from src") + createPartition("2010-01-02", 11, + "SELECT '1', 'A' from src UNION ALL SELECT '1', 'A' from src") + + sql(s"ANALYZE TABLE $tableName PARTITION (ds, hr) COMPUTE STATISTICS NOSCAN") + + assertPartitionStats("2010-01-01", "10", rowCount = None, sizeInBytes = 2000) + assertPartitionStats("2010-01-01", "11", rowCount = None, sizeInBytes = 2000) + assertPartitionStats("2010-01-01", "11", rowCount = None, sizeInBytes = 2000) + assertPartitionStats("2010-01-02", "11", rowCount = None, sizeInBytes = 2*2000) + + sql(s"ANALYZE TABLE $tableName PARTITION (ds, hr) COMPUTE STATISTICS") + + assertPartitionStats("2010-01-01", "10", rowCount = Some(500), sizeInBytes = 2000) + assertPartitionStats("2010-01-01", "11", rowCount = Some(500), sizeInBytes = 2000) + assertPartitionStats("2010-01-01", "11", rowCount = Some(500), sizeInBytes = 2000) + assertPartitionStats("2010-01-02", "11", rowCount = Some(2*500), sizeInBytes = 2*2000) + } + } + + test("analyze partitions for an empty table") { + val tableName = "analyzeTable_part" + + withTable(tableName) { + sql(s"CREATE TABLE $tableName (key STRING, value STRING) PARTITIONED BY (ds STRING)") + + // make sure there is no exception + sql(s"ANALYZE TABLE $tableName PARTITION (ds) COMPUTE STATISTICS NOSCAN") + + // make sure there is no exception + sql(s"ANALYZE TABLE $tableName PARTITION (ds) COMPUTE STATISTICS") + } + } + + test("analyze non-existent partition") { + val tableName = "analyzeTable_part" + withTable(tableName) { + sql(s"CREATE TABLE $tableName (key STRING, value STRING) PARTITIONED BY (ds STRING)") + + sql(s"INSERT INTO TABLE $tableName PARTITION (ds='2010-01-01') SELECT * FROM src") + + intercept[AnalysisException] { + sql(s"ANALYZE TABLE $tableName PARTITION (hour=20) COMPUTE STATISTICS") + } + + intercept[AnalysisException] { --- End diff -- Could you capture all these error messages? We always need to check whether the error messages make sense. ```Scala val message = intercept[AnalysisException] { ... }.getMessage assert(message.contains("XYZ")) ```
--- 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