zhong.zhu created KYLIN-5740:
--------------------------------
Summary: When the query includes multiple aggregate functions, and
the measure corresponding to the aggregate functions is not in the aggregate
index, the query may report an error
Key: KYLIN-5740
URL: https://issues.apache.org/jira/browse/KYLIN-5740
Project: Kylin
Issue Type: Bug
Affects Versions: 5.0-beta
Reporter: zhong.zhu
Assignee: zhong.zhu
Fix For: 5.0.0
Attachments: image-2023-12-11-14-25-14-409.png
Reproduction method:
Table Building Statements
{code:sql}
CREATE EXTERNAL TABLE LT859 (
`PRO_MT_SEASON` varchar(4096),
`ARRIVAL_TRANSIT_IN_PRM_AMT` decimal(18,4),
`BIG_REGION_NAME` varchar(4096),
`REGION_NAME` varchar(4096)
) ROW FORMAT DELIMITED FIELDS TERMINATED BY ','
STORED AS TEXTFILE;LOAD DATA INPATH '/path/to/data/LT859.csv' OVERWRITE INTO
TABLE LT859;
{code}
Data（LT859.csv）
{code:bash}
22Q1,150.12,东区,华东区
22Q2,200.59,中区,华中区
22Q3,301.55,上海区,华东区
{code}
Select SQL:
{code:sql}
select
max(BIG_REGION_NAME) "max1",
max(
case when 1=1 then "BIG_REGION_NAME"
end
) "max2"
from XXX.LT859
{code}
Create model, dimension 4 columns selected, add metrics MAX(BIG_REGION_NAME)
Save the model and delete the base aggregation index, manually add an
aggregation group as a replacement for the base aggregation, but it can't have
the MAX(BIG_REGION_NAME) metric in it to mimic the dim as measure scenario, and
build the full amount.
Using the above select sql query, you can see that NPE is reported in the
background:
!image-2023-12-11-14-25-14-409.png!
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
