Data types for Kafka connector 
 When reading data using the Kafka table connector, you must specify the format of the
                incoming messages so that Flink can map incoming data to table columns
                properly.
JSON format  The JSON format enables you to read and write JSON data. You must add the JSON         dependency to your project and define the format type in CREATE table to JSON.CSV format  The CSV format allows your applications to read data from, and write data to different       external sources in CSV. You must add the CSV dependency to your project and define the format       type in CREATE table to CSV.Avro format  The Apache Avro format enables you to read and write Avro data. You must add the Avro     dependency to your project and define the format type in CREATE table to Avro. You also need to     specify the fields of the Avro record within the table.Schema Registry formats  You can avoid defining the Avro and JSON schema for Kafka table sources and sinks, when     the schema is stored in Cloudera Schema Registry.