You have a DataFrame df with columns order_id, customer_id, and product_id. How would you validate that order_id is unique across all rows using Spark? Provide the code snippet.