[
https://issues.apache.org/jira/browse/HIVE-28301?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sercan Tekin updated HIVE-28301:
--------------------------------
Description:
*Create a table and insert data into it:*
{code:java}
CREATE TABLE tbl_1 (col_1 STRING);
INSERT INTO tbl_1 VALUES ('G'),('G'),('not G'),('G'),('G'),('G');
{code}
*Submit the below query:*
{code:java}
SELECT DISTINCT (
CASE
WHEN col_1 = "G" THEN "Value_1"
WHEN substr(LPAD(col_1,3,"0") ,1,1) = "G" THEN "Value_2"
ELSE "Value_3"
END) AS G
FROM tbl_1;
{code}
*Actual result:*
{code:java}
3alue_1
Value_1
Value_3
nValue_
{code}
*Expected result (This is what Hive-2.3 returns):*
{code:java}
Value_1
Value_3
{code}
*Workaround:*
Either disabling vectorization;
{code:java}
SET hive.vectorized.execution.enabled=false;
{code}
Or reverting https://issues.apache.org/jira/browse/HIVE-16731.
CC: [~teddy.choi] and [~mmccline] as HIVE-16731 was reported and fixed by you
guys.
was:
*Create a table and insert data into it:*
{code:java}
CREATE TABLE tbl_1 (col_1 STRING);
INSERT INTO tbl_1 VALUES ('G'),('G'),('not G'),('G'),('G'),('G');
{code}
*Submit the below query:*
{code:java}
SELECT DISTINCT (
CASE
WHEN col_1 = "G" THEN "Value_1"
WHEN substr(LPAD(col_1,3,"0") ,1,1) = "G" THEN "Value_2"
ELSE "Value_3"
END) AS G
FROM tbl_1;
{code}
*Actual result:*
{code:java}
3alue_1
Value_1
Value_3
nValue_
{code}
*Expected result (This is what Hive-2.3 returns):*
{code:java}
Value_1
Value_3
{code}
*Workaround:*
Either disabling vectorization;
{code:java}
SET hive.vectorized.execution.enabled=false;
{code}
Or reverting https://issues.apache.org/jira/browse/HIVE-16731.
> Vectorization: CASE WHEN Returns Wrong Result in Hive-3.1.3
> -----------------------------------------------------------
>
> Key: HIVE-28301
> URL: https://issues.apache.org/jira/browse/HIVE-28301
> Project: Hive
> Issue Type: Bug
> Components: Vectorization
> Affects Versions: 3.1.3
> Reporter: Sercan Tekin
> Priority: Critical
>
> *Create a table and insert data into it:*
> {code:java}
> CREATE TABLE tbl_1 (col_1 STRING);
> INSERT INTO tbl_1 VALUES ('G'),('G'),('not G'),('G'),('G'),('G');
> {code}
> *Submit the below query:*
> {code:java}
> SELECT DISTINCT (
> CASE
> WHEN col_1 = "G" THEN "Value_1"
> WHEN substr(LPAD(col_1,3,"0") ,1,1) = "G" THEN "Value_2"
> ELSE "Value_3"
> END) AS G
> FROM tbl_1;
> {code}
> *Actual result:*
> {code:java}
> 3alue_1
> Value_1
> Value_3
> nValue_
> {code}
> *Expected result (This is what Hive-2.3 returns):*
> {code:java}
> Value_1
> Value_3
> {code}
> *Workaround:*
> Either disabling vectorization;
> {code:java}
> SET hive.vectorized.execution.enabled=false;
> {code}
> Or reverting https://issues.apache.org/jira/browse/HIVE-16731.
> CC: [~teddy.choi] and [~mmccline] as HIVE-16731 was reported and fixed by you
> guys.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
