Details
-
Bug
-
Status: Resolved
-
Major
-
Resolution: Fixed
-
None
-
None
Description
Currently, user can insert data even if there is small columns compare to target columns as follows:
create table table1 (id int, name text, score float, type text) using csv with ('csvfile.delimiter'='|') ; create table table3 (id int, name text, score float, type text) using csv with ('csvfile.delimiter'='|') ; create table table4 (id int, name text, score float, type text) using csv with ('csvfile.delimiter'='|') partition by column(part text); default> insert overwrite into table3 select id, name from table1; Progress: 0%, response time: 0.544 sec Progress: 100%, response time: 1.158 sec (5 rows, 1.158 sec, 60 B inserted) default> insert overwrite into table4 select id, name from table1; Progress: 100%, response time: 0.468 sec (5 rows, 0.468 sec, 65 B inserted)
As above, tajo can run a query for inserting data. But in partitioned table, tajo made null value for partition column as follows:
hdfs dfs -ls /tajo/warehouse/default/table4 Found 1 items drwxr-xr-x - blrunner supergroup 0 2014-05-26 16:27 /tajo/warehouse/default/table4/part=
Thus we need to make a syntax exception for blocking null partition value.