Hi Bert-Jan,
I see, I checked the first stack trace, which I believe to be an issue within the JDBC driver. The second one (which I had overlooked in my previous email, sorry) is also a problem of the driver in my opinion. I don't think a type name should be nullable. But we can work around that easily, indeed, we'll just translate it to SQLDataType.OTHER. I've created an issue for this, which will be backported to 3.21.3, 3.20.14, and 3.19.33:
You can test the fix by applying this patch to your 3.21.1 version:
Or you can wait for the nightly snapshot builds to build.
Let me know if this addresses the issue for you.
Best Regards,
Lukas