cloud-fan commented on pull request #30521:
URL: https://github.com/apache/spark/pull/30521#issuecomment-737038157


   `DataFrameWriterV2` is very powerful to describe the table writing behavior 
(CREATE, CREATE IF NOT EXISTS, CREATE OR REPLACE, REPLACE, append, overwrite 
where, etc.) and I don't think the current streaming framework can support 
these at the current stage.
   
   Ideally we need to handle these cases:
   1. table exists and users want to write to it
   2. table not exists and users want to fail
   3. table not exists and users want to create it
   
   The current PR can't cover case 2 but I don't know how common it is for 
streaming users. Adding a `DataStreamWriterV2` to  cover case 2 looks an 
overkill to me. One possible solution is to add 2 methods `insertTable` and 
`createAndInsertTable`. If we think case 2 is rare, adding only `toTable` which 
works as `createAndInsertTable` is also fine to me.


----------------------------------------------------------------
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