Pengcheng Xiong created HIVE-16169:
--------------------------------------
Summary: Improve StatsOptimizer to deal with groupby partition
columns
Key: HIVE-16169
URL: https://issues.apache.org/jira/browse/HIVE-16169
Project: Hive
Issue Type: Bug
Reporter: Pengcheng Xiong
Assignee: Pengcheng Xiong
As reported by [~ashutoshc]
1) select sum(c), count(c),... from T group by b;
2) select max(c), min(c), ... from T group by b;
If b happens to be a partition column, we can also answer these from metadata.
Currently, StatsOptimizer don't handle these queries, but we can extend it to
handle those as well.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
