Writing large spark data frame as parquet to s3 bucket

Question

My Scenario

• I have a spark data frame in a AWS glue job with 4 million records
• I need to write it as a SINGLE parquet file in AWS s3