Hello,
I am attempting to create a data type, in ER Studio, that will be forward-engineered into an emerging cloud DW database platform called 'SnowFlake', which is ANSI-compliant, except that they have a proprietary data type called 'Variant' to store my JSON (as well as other non-relational types). The people at Snowflake (www.snowflakecomputing.com) say that the forward-engineered DDL code need only specify (null / not null and) "Variant" as datatype (no length, precision, or other items), and that DDL can be executed just like that in Snowflake.
However, on my first attempt, using ER Studio logical model I created, I created a User-Defined Datatype, which I named 'Variant'. Being forced to choose a base datatype, I chose Char(10) (temporarily thinking I would need to change that, but not knowing how). So, I then created a "JSON" domain to use the new UDT named 'Variant', and bound a JSON-containing field to that domain, verifying that the data type 'Variant' now showed up as the field's type. Now the good/bad news...
Good (but only temporarily): Upon generating physical model (choosing 'Generic DBMS') from logical, the new physical model did show the 'Variant' datatype in my chosen field, which temporarily made me think I had done it.
Not Good: Next, I clicked on 'DDL', and noticed that, sadly, the chosen field had the Char (10) data type for purposes of DDL.
Help Requested: So, I request help in understanding how to actually create a UDT that can be generated to a physical model, then create the DDL with a field bound as described above, so that it can be forward-engineered with this new data type 'Variant'.
What to do?