[
https://issues.apache.org/jira/browse/CARBONDATA-1203?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16060325#comment-16060325 ]
chenerlu edited comment on CARBONDATA-1203 at 6/23/17 2:36 AM:
---------------------------------------------------------------
Hi, I encounter same problem. Issue can be summarized as follows.
Step 1: create a carbon table.
cc.sql("CREATE TABLE IF NOT EXISTS t3 (id Int, name String) STORED BY 'carbondata'")
Step 2: load data, then t3 will have 10 records
cc.sql("LOAD DATA LOCAL INPATH 'mypathofdata' INTO TABLE t3 ")
Step 3: insert constant into table t3
cc.sql("INSERT INTO TABLE t3 SELECT 1, 'jack' FROM t3")
Step4: count table t3
cc.sql("SELECT count (\*) FROM t3")
Actual result: t3 will have 20 records. (20 records = 10 + 10, the second '10' is because t3 has 10 records, if we change t3 to t4 which have 5 records, the result will be 15, so I think carbondata handle constant as '\*', not sure, this should be confirm).
Expected result: t3 should have 11 records or throw sql.AnalysisException (This will be same as Hive table I think)
Any idea about this issue, which solution is better ?
[~ravi.pesala] [~chenliang613]
was (Author: chenerlu):
Hi, I encounter same problem. Issue can be summarized as follows.
Step 1: create a carbon table.
cc.sql("CREATE TABLE IF NOT EXISTS t3 (id Int, name String) STORED BY 'carbondata'")
Step 2: load data, then t3 will have 10 records
cc.sql("LOAD DATA LOCAL INPATH 'mypathofdata' INTO TABLE t3 ")
Step 3: insert constant into table t3
cc.sql("INSERT INTO TABLE t3 SELECT 1, 'jack' FROM t3")
Step4: count table t3
cc.sql("SELECT count (\*) FROM t3")
Actual result: t3 will have 20 records.
Expected result: t3 should have 11 records or throw sql.AnalysisException (This will be same as Hive table I think)
Any idea about this issue, which solution is better ?
[~ravi.pesala] [~chenliang613]
> insert data caused many duplicated data on spark 1.6.2
> -------------------------------------------------------
>
> Key: CARBONDATA-1203
> URL:
https://issues.apache.org/jira/browse/CARBONDATA-1203> Project: CarbonData
> Issue Type: Bug
> Reporter: Jarck
>
> I use branch-1.1 do insert test on spark 1.6.2 in my local machine
> I try to run the sql as below to insert a data
> spark.sql(s"""
> insert into $tableName select $id,'$date','$country','$testName'
> ,'$phoneType','$serialname',$salary from $tableName
> """).show()
> at last the data has been inserted successfully, but it inserted many duplicated data
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
Locked
