How to build highly scalable Data Persistence Pipeline with AWS S3 Sink Connector