alexmc-ms commented on issue #10713: URL: https://github.com/apache/airflow/issues/10713#issuecomment-698097781
So this is now an official stance? There is no support for MS SQL Server in Airflow? As someone who has used SQL Server with Airflow for nearly a year now, this is disappointing. Besides this new bit issue, I've only encountered one real problem: SQL Server can't use an index if the string type VARCHAR vs NVARCHAR is mismatched. Since it uses pyodbc, all strings from sqlalchemy are unicode. This requires table creation for any strings used in an index as Unicode rather than String. This isn't optimal, but it's pretty easy to patch since table creation is a one time thing. And this same patch can be applied to any new migrations across Airflow updates. In my mind, there is a distinction between saying something is "tested with" versus saying something is not supported such that if there is a regression, it won't be fixed. Even as recently as 1.10.10 a SQL Server regression was fixed (https://github.com/apache/airflow/pull/8385). This led me to believe it is still supported. This, and what jmfreeman mentioned above. We are weeks away from shipping to thousands of customers. If SQL Server is truly dead in Airflow, I suppose my long days are about to get even longer. ---------------------------------------------------------------- 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]
