bucket table

[lionel061201]

Hi Dev,
I created a bucket table and loaded 100 million rows of data. But there is
only one data file and one index file in path.
It returned a "partially success" error when loading was done.

I debugged into the program and it shows bad record is not null.
Any one can explain why bad record will cause incorrect bucket number and
is there any prompt for users to fix the bad record issue?

Thanks,
Lionel

[David CaiQiang]

Hi Lu,

Please find faq page: http://carbondata.apache.org/docs/latest/faq.html

I think you should add option 'BAD_RECORDS_ACTION'='FORCE' to  load sql.
e.g. load data ... options( 'BAD_RECORDS_ACTION'='FORCE')

Best Regards
David QiangCai