You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@carbondata.apache.org by "dhatchayani (JIRA)" <ji...@apache.org> on 2017/11/15 12:44:00 UTC

[jira] [Created] (CARBONDATA-1724) Wrong data displayed for <= filter for timestamp column(dictionary column)

dhatchayani created CARBONDATA-1724:
---------------------------------------

             Summary: Wrong data displayed for <= filter for timestamp column(dictionary column)
                 Key: CARBONDATA-1724
                 URL: https://issues.apache.org/jira/browse/CARBONDATA-1724
             Project: CarbonData
          Issue Type: Bug
            Reporter: dhatchayani
            Assignee: dhatchayani


*Issue:*
<= filter is giving wrong results for timestamp dictioinary column

*Steps to reproduce:*
(1) Create a table with a timestamp dictionary column
create table t1(dob timestamp) stored by 'carbondata' TBLPROPERTIES ('DICTIONARY_INCLUDE'='dob')
(2) Load data
1970-01-01 05:30:00 (same value as 300 records)
(3) Apply filter on table
select count(*) from t1 where dob=cast('1970-01-01 05:30:00' as timestamp);
+--------+
|count(1)|
+--------+
|300     |
+--------+
select count(*) from t1 where dob<=cast('1970-01-01 05:30:00' as timestamp);
+--------+
|count(1)|
+--------+
|1         |
+--------+


Both the queries should give us the same results.

Solution:
In less than equal to filter, we are considering surrogate 2 as default value. But surrogate 1 is for default value.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)