SQL Runner: dummy columns in Spark tables

Hi guys!

I ran into a problem while fetching the schema of a Spark table in SQL Runner. As I can see, Looker is using the following SQL command for this.

DESCRIBE <db>.<table>

And this command produces a result in this manner:

col_name data_type comment
my_string_column string
my_date_column date
my_int_column bigint
# Partitioning
Part 0
my_partition_key string

Which results in appearing of dummy columns in the parsed schema

Does anybody know any workarounds in such case?

This is planned to be addressed in 21.6

1 Like