Union All With List Datatype

Hi, I am finding an issue when trying to combine two virtual datasets using UNION ALL. When both column datatypes are a ‘list’ type I am getting back a Column Mismatch Error. This is a similar issue that I have when I use the coalesce() function on those same two columns.

Any suggestions on how to resolve this?


Do the elements inside the LIST datatype have the same datatype?


Thanks for the reply. The elements inside the list are a complex nested array, and I have been evaluating their schema using the job api. The schema’s vary by a few fields. So to that my next questions would be, 1) Do the underlying schemas have to be the same for this to work and if so 2) Does Dremio allow you to change the underlying schema of a field directly?


Dremio will do schema learning, max 10 retries per run, but you can keep trying, Dremio does not allow change of a field, has to be done at the source