Github user wzhfy commented on a diff in the pull request:
https://github.com/apache/spark/pull/18421#discussion_r126273773
--- Diff:
sql/hive/src/test/scala/org/apache/spark/sql/hive/StatisticsSuite.scala ---
@@ -201,6 +202,193 @@ 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
+ }
+
+ 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")
+ sql(
+ s"""
+ |INSERT INTO TABLE $tableName PARTITION (ds='2010-01-02')
+ |SELECT * FROM src
+ |UNION ALL
+ |SELECT * FROM src
+ """.stripMargin)
+ sql(s"INSERT INTO TABLE $tableName PARTITION (ds='2010-01-03')
SELECT * FROM src")
+
+ sql(s"ANALYZE TABLE $tableName PARTITION (ds='2010-01-01') COMPUTE
STATISTICS").collect()
+
+ sql(s"ANALYZE TABLE $tableName PARTITION (ds='2010-01-02') COMPUTE
STATISTICS").collect()
+
+ assert(queryStats("2010-01-01").rowCount.get === 500)
+ assert(queryStats("2010-01-01").sizeInBytes === 5812)
+
+ assert(queryStats("2010-01-02").rowCount.get === 2*500)
+ assert(queryStats("2010-01-02").sizeInBytes === 2*5812)
+ }
+ }
+
+ test("analyze single partition noscan") {
+ val tableName = "analyzeTable_part"
+
+ def queryStats(ds: String): CatalogStatistics = {
+ val partition =
+
spark.sessionState.catalog.getPartition(TableIdentifier(tableName), Map("ds" ->
ds))
+ partition.stats.get
+ }
+
+ 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")
+ sql(
+ s"""
+ |INSERT INTO TABLE $tableName PARTITION (ds='2010-01-02')
+ |SELECT * FROM src
+ |UNION ALL
+ |SELECT * FROM src
+ """.stripMargin)
+ sql(s"INSERT INTO TABLE $tableName PARTITION (ds='2010-01-03')
SELECT * FROM src")
+
+ sql(s"ANALYZE TABLE $tableName PARTITION (ds='2010-01-01') COMPUTE
STATISTICS NOSCAN")
+ .collect()
+
+ sql(s"ANALYZE TABLE $tableName PARTITION (ds='2010-01-02') COMPUTE
STATISTICS NOSCAN")
+ .collect()
+
+ assert(queryStats("2010-01-01").rowCount === None)
+ assert(queryStats("2010-01-01").sizeInBytes === 5812)
+
+ assert(queryStats("2010-01-02").rowCount === None)
+ assert(queryStats("2010-01-02").sizeInBytes === 2*5812)
+ }
+ }
+
+ test("analyze a set of partitions") {
--- End diff --
@mbasmanova IIUC, the logic is wrong here. For example, when analyzing
partition (ds, hr), we should not remove them in parser. Currently we parse it
to `AnalyzeTableCommand`, which collects table-level stats. But what we need to
do is to collect **partition-level stats for all partitions**.
Check hive's behavior
[here](https://cwiki.apache.org/confluence/display/Hive/StatsDev#StatsDev-Examples)
---
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 [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
