Here are some troubleshooting observed actions that seem to have helped with this "WriteToDatabaseTableFrom_TransformForOutputToDatabaseTableFrom" type of message error:
1. Every time you open Dataflow Gen2 to edit the dataflow query or data destination etc, you should redo the connection details eg click edit connection and go through steps to choose your connection. Maybe not, but seems to me it might forget the connection?
2. Ensure your dataflow query results are all assigned data types.
* Always use datetime not date even if your date is just a date (Delta table feature!)
* Always assign data type to each column. If you forget to do assign , dataflow assigns type "Any" to the column which cannot be written to a table. The data destination schema page will give you a message to this effect.
* Some steps in query may make a column "forget" its data type if it was assigned in previous steps, so it seems you should ensure that final step assigns all columns their data type. This is the same in Dataflow "gen 1" too. Power BI and Excel queries are a lot more forgiving about this eg they will create the data results and you can use them without any errors (although their data type might not work in relationships or as expected in results and visuals).
3. If you change columns or their data types in your dataflow query then you should delete the data destination table from the Lakehouse (and DWH but haven't worked with DWH tables yet) and then run the updated dataflow again to recreate a new updated table. It doesn't appear to update an existing table schema eg add new column or remove or modify existing columns.
4. Learn reason dataflow failed. In the Workspace list of objects eg dataflows, lakehouse, report, dataset etc if the dataflow fails there will be red triangle. You can click on this, or click on the 3 dots and open Refresh history and 1) click on the download symbole to get the csv export and see some info on this. Or better let, 2) click on the history date time link to see more details. This is where I see the "WriteToDatabaseTableFrom_TransformForOutputToDatabaseTableFrom" message.
5. Also refreshing any of the lists of items and also refreshing the entire browser page really helps to show all most recent status and objects. For example, sometimes a newly added table doesn't show until page is refreshed.
6. Also make sure that the data destination table selection for new or existing is correctly set. If you create a new dataflow and set to new, then subsequent runs seem to change it to existing. But when you delete the table so you can create a new table with new columns or modified type columns, then you need to switch it back to new.
7. Do not ever delete the Lakehouse or DWH staging artifacts. These should probably be hidden or hideable. They seem to collect historical content eg retaining previously deleted tables but in a way that each one is unique item. Not sure what happens if it has too much historical data or if that will ever be a problem.
This "WriteToDatabaseTableFrom_TransformForOutputToDatabaseTableFrom" type of message error message actually seems quite informative basically that something is preventing dataflow results from being written to whereever you are putting it eg in Lakehouse or DWH.
