[CARBONDATA-2085] It's different between load twice and create datamap with load again after load data and create datamap

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.3.0
• Fix Version/s: 1.4.0
• Component/s: core, spark-integration
• Assignee: Unassigned
• Reporter: Bo Xu
• Created: 26/Jan/18 03:37
• Updated: 11/Feb/18 06:15
• Resolved: 11/Feb/18 06:15

Description

It's different between two test case

test case 1: load twice and create datamap , and then query
test case 2:load once , create datamap and load again, and then query

+  test("load data into mainTable after create timeseries datamap on table 1") {
 +    sql("drop table if exists mainTable")
 +    sql(
 +      """
 +        | CREATE TABLE mainTable(
 +        |   mytime timestamp,
 +        |   name string,
 +        |   age int)
 +        | STORED BY 'org.apache.carbondata.format'
 +      """.stripMargin)
 +
 +    sql(s"LOAD DATA LOCAL INPATH '$resourcesPath/timeseriestest.csv' into table mainTable")
 +
 +    sql(
 +      """
 +        | create datamap agg0 on table mainTable
 +        | using 'preaggregate'
 +        | DMPROPERTIES (
 +        |   'timeseries.eventTime'='mytime',
 +        |   'timeseries.hierarchy'='second=1,minute=1,hour=1,day=1,month=1,year=1')
 +        | as select mytime, sum(age)
 +        | from mainTable
 +        | group by mytime""".stripMargin)
 +
 +    sql(s"LOAD DATA LOCAL INPATH '$resourcesPath/timeseriestest.csv' into table mainTable")
 +    val df = sql(
 +      """
 +        | select
 +        |   timeseries(mytime,'minute') as minuteLevel,
 +        |   sum(age) as sum
 +        | from mainTable
 +        | where timeseries(mytime,'minute')>='2016-02-23 01:01:00'
 +        | group by
 +        |   timeseries(mytime,'minute')
 +        | order by
 +        |   timeseries(mytime,'minute')
 +      """.stripMargin)
 +
 +    // only for test, it need remove before merge
 +    df.show()
 +    sql("select * from maintable_agg0_minute").show(100)
 +
 +    checkAnswer(df,
 +      Seq(Row(Timestamp.valueOf("2016-02-23 01:01:00"), 120),
 +        Row(Timestamp.valueOf("2016-02-23 01:02:00"), 280)))
 +
 +  }
 +
 +  test("load data into mainTable after create timeseries datamap on table 2") {
 +    sql("drop table if exists mainTable")
 +    sql(
 +      """
 +        | CREATE TABLE mainTable(
 +        |   mytime timestamp,
 +        |   name string,
 +        |   age int)
 +        | STORED BY 'org.apache.carbondata.format'
 +      """.stripMargin)
 +
 +    sql(s"LOAD DATA LOCAL INPATH '$resourcesPath/timeseriestest.csv' into table mainTable")
 +    sql(s"LOAD DATA LOCAL INPATH '$resourcesPath/timeseriestest.csv' into table mainTable")
 +    sql(
 +      """
 +        | create datamap agg0 on table mainTable
 +        | using 'preaggregate'
 +        | DMPROPERTIES (
 +        |   'timeseries.eventTime'='mytime',
 +        |   'timeseries.hierarchy'='second=1,minute=1,hour=1,day=1,month=1,year=1')
 +        | as select mytime, sum(age)
 +        | from mainTable
 +        | group by mytime""".stripMargin)
 +
 +
 +    val df = sql(
 +      """
 +        | select
 +        |   timeseries(mytime,'minute') as minuteLevel,
 +        |   sum(age) as sum
 +        | from mainTable
 +        | where timeseries(mytime,'minute')>='2016-02-23 01:01:00'
 +        | group by
 +        |   timeseries(mytime,'minute')
 +        | order by
 +        |   timeseries(mytime,'minute')
 +      """.stripMargin)
 +
 +    // only for test, it need remove before merge
 +    df.show()
 +    sql("select * from maintable_agg0_minute").show(100)
 +
 +
 +    checkAnswer(df,
 +      Seq(Row(Timestamp.valueOf("2016-02-23 01:01:00"), 120),
 +        Row(Timestamp.valueOf("2016-02-23 01:02:00"), 280)))
 +  }
 +

test case 1 and 2 should success , but test case 1 fail

Issue Links

• links to: GitHub Pull Request #1870