Spark: rdd.count() and rdd.write() are executing transformations twice

Question

I am using Apache Spark to fetch records from database and after some transformations, writing them to AWS S3. Now I also want to count the no of records I am writing to S3