Naresh P R created HIVE-25084: --------------------------------- Summary: Incorrect aggregate results on bucketed table Key: HIVE-25084 URL: https://issues.apache.org/jira/browse/HIVE-25084 Project: Hive Issue Type: Bug Reporter: Naresh P R
Steps to repro {code:java} CREATE TABLE test_table( col1 int, col2 char(32), col3 varchar(3)) CLUSTERED BY (col2) SORTED BY ( col2 ASC, col3 ASC, col1 ASC) INTO 32 BUCKETS stored as orc; set hive.query.results.cache.enabled=false; insert into test_table values(2, "123456", "15"); insert into test_table values(1, "123456", "15"); SELECT col2, col3, max(col1) AS max_sequence FROM test_table GROUP BY col2, col3; ==> LocalFetch correct result <== 123456 15 2 ==> Wrong result with Tez/Llap <== set hive.fetch.task.conversion=none; 123456 15 2 123456 15 1 ==> Correct result with Tez/Llap disabling map aggregation <== set hive.map.aggr=false; 123456 15 2 {code} -- This message was sent by Atlassian Jira (v8.3.4#803005)