ayushtkn commented on PR #4250:
URL: https://github.com/apache/hive/pull/4250#issuecomment-1515410864

   I won't say we should go on like that unless we have a reason. There is a 
recent ticket: [HIVE-25965](https://issues.apache.org/jira/browse/HIVE-25965), 
The test passes on flaky job, passes locally and only fails when other tests 
are run together, because plan changes as the state of metastore changes & that 
actually highlighted a production bug, what we actually want from a Unit Test, 
**break before things break in actual prod**
   
   [HIVE-25965](https://issues.apache.org/jira/browse/HIVE-25965) is the reason 
I along with many others were screwed downstream for more than a week, because 
that test was failing on gerrit but not on flaky job neither locally.
   
   Shouldn't encourage this approach which can destabilise the build. No harm 
waiting and figuring out **Why it failed, It may be a prod bug and we can fix 
it**
   
   
   Just to heighlight, the test initially failed claiming **wrong data**. 
Instead of returning the inserted value 1.5 it returned 1.0
   
   ```
   
   < 1.0
   --
   ---
   > 1.5
   
   ```
   
   It wasn't some random connection issue or timeout issue, Wrong data itself 
is very serious stuff, we should figure out the cause for that
   
   Failure Trace:
   
https://issues.apache.org/jira/secure/attachment/13000564/TestMiniLlapLocalCliDriver.testCliDriver%5Bschema_evol_par_vec_table_non_dictionary_encoding%5D.txt
   
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to