[
https://issues.apache.org/jira/browse/CARBONDATA-2351?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
xhnqccf updated CARBONDATA-2351:
--------------------------------
Description:
Carbondata1.3.1 with Spark2.1.2,after insert into values,SELECT is right.but
next day,data of SELECT is null but the line number is right.
create table carbon01(id int,name int,age int,sex int) stored by 'carbondata';
insert into carbon01 values(1,1,1,1);
select * from carbon01;
1 1 1 1
then,i exit spark-sql.
but,next day,
select * from carbon01;
null null null null
.it may not because of time.
after one line turn to null,i insert into again.
insert into carbon01 values(2,2,2,2);
then i select,the result is:
null null null null
2 2 2 2
the log is:
ResultCollectorFactory:[Executor task launch worker for task 1][partition...]
Restructure dictionary vector collector is used to scan and collect the data
ResultCollectorFactory:[Executor task launch worker for task 2][partition...]
Vector based dictionary collector is used to scan and collect the data
and i follow source code,in task1,
tableBlockInfo.getDetailInfo.columnSchemas.get(0).isDimensionColumn=true;
in task2,
tableBlockInfo.getDetailInfo.columnSchemas.get(0).isDimensionColumn=false;
was:
Carbondata1.3.1 with Spark2.1.2,after insert into values,SELECT is right.but
next day,data of SELECT is null but the line number is right.
create table carbon01(id int,name int,age int,sex int) stored by 'carbondata';
insert into carbon01 values(1,1,1,1);
select * from carbon01;
1 1 1 1
then,i exit spark-sql.
but,next day,
select * from carbon01;
null null null null
> CarbonData Select null
> ----------------------
>
> Key: CARBONDATA-2351
> URL: https://issues.apache.org/jira/browse/CARBONDATA-2351
> Project: CarbonData
> Issue Type: Bug
> Affects Versions: 1.3.1
> Environment: Carbondata1.3.1 with Spark2.1.2
> Reporter: xhnqccf
> Priority: Major
> Labels: SELECT, null
>
> Carbondata1.3.1 with Spark2.1.2,after insert into values,SELECT is right.but
> next day,data of SELECT is null but the line number is right.
> create table carbon01(id int,name int,age int,sex int) stored by 'carbondata';
> insert into carbon01 values(1,1,1,1);
> select * from carbon01;
> 1 1 1 1
> then,i exit spark-sql.
> but,next day,
> select * from carbon01;
> null null null null
> .it may not because of time.
> after one line turn to null,i insert into again.
> insert into carbon01 values(2,2,2,2);
> then i select,the result is:
> null null null null
> 2 2 2 2
> the log is:
> ResultCollectorFactory:[Executor task launch worker for task 1][partition...]
> Restructure dictionary vector collector is used to scan and collect the data
> ResultCollectorFactory:[Executor task launch worker for task 2][partition...]
> Vector based dictionary collector is used to scan and collect the data
> and i follow source code,in task1,
> tableBlockInfo.getDetailInfo.columnSchemas.get(0).isDimensionColumn=true;
> in task2,
> tableBlockInfo.getDetailInfo.columnSchemas.get(0).isDimensionColumn=false;
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
