Currently we're identifying if a column is virtual in Snowflake by running a SHOW COLUMN query and checking the KIND field for VIRTUAL_COLUMN. Unfortunately, there's a 10k limit on entries returned from SHOW queries in Snowflake and we'd like to be able to run this query at the schema level on schemas ~25k tables.

According to this post there's no way to identify virtual columns in the information_schema.columns view and we'd like to avoid having to run a SHOW COLUMNS query at the table level or having to run a desc table on every table.

Is there some other way we can identify virtual columns at scale?