[ 
https://issues.apache.org/jira/browse/HIVE-22357?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Krishnadevan Purushothaman updated HIVE-22357:
----------------------------------------------
           Attachment: Hive-JDBC schema matching sensitivity.txt
          Component/s: HiveServer2
                       Hive
    Affects Version/s: 3.1.0
          Description: 
The problem is that, for certain types of schema mismatch between the Hive 
table definition and the "hive.sql.query" parameter, Hive simply returns no 
rows rather than throwing an error when queried.

Ideally Hive would check for schema compatibility during table definition and 
throw an error if there's a problem - the earlier the better. But even if that 
cannot be made a requirement, I'd definitely expect an error rather than a 
silent failure (i.e. zero rows returned) at runtime.

I'm attaching investigation of this issue in "Hive-JDBC schema matching 
sensitivity.txt".
             Priority: Critical  (was: Major)
              Summary: Schema mismatch between the Hive table definition and 
the "hive.sql.query" Parameter  (was: Schema mismatch between the Hive table 
definition and the "hive.sql.query)

> Schema mismatch between the Hive table definition and the "hive.sql.query" 
> Parameter
> ------------------------------------------------------------------------------------
>
>                 Key: HIVE-22357
>                 URL: https://issues.apache.org/jira/browse/HIVE-22357
>             Project: Hive
>          Issue Type: Improvement
>          Components: Hive, HiveServer2
>    Affects Versions: 3.1.0
>            Reporter: Krishnadevan Purushothaman
>            Priority: Critical
>         Attachments: Hive-JDBC schema matching sensitivity.txt
>
>
> The problem is that, for certain types of schema mismatch between the Hive 
> table definition and the "hive.sql.query" parameter, Hive simply returns no 
> rows rather than throwing an error when queried.
> Ideally Hive would check for schema compatibility during table definition and 
> throw an error if there's a problem - the earlier the better. But even if 
> that cannot be made a requirement, I'd definitely expect an error rather than 
> a silent failure (i.e. zero rows returned) at runtime.
> I'm attaching investigation of this issue in "Hive-JDBC schema matching 
> sensitivity.txt".



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

Reply via email to