yaooqinn commented on pull request #28963:
URL: https://github.com/apache/spark/pull/28963#issuecomment-652756928


   > What's the behavior before and after this PR? I thought this PR is only to 
pick the actual exception, not add/remove stacktraces.
   
   Yes, we do not change the exception itself but only pick things to deliver 
to the JDBC-client side through error responses. , which eventually result in 
the `reason` filed of ` java.sql.SQLException` at client-side.
   
   Before this PR and w/ SPARK-29283, only the error messages of the root 
causes will be delivered. In this case, many intentionally made exceptions will 
be omitted, e.g. `SparkUpgradeException`, `AnalysisException`.  Thus, the root 
may not always be the actual one.
   After this PR, we fill the error response with the full stacktraces like 
Hive.
   
   WDYT about adding a configuration to control this behavior? @cloud-fan. 
Personally, I'm -1 of it, because the change is trivial and not really a 
behavior change and it is consistent with hive currently.
   
   


----------------------------------------------------------------
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.

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