Hi, I am trying to author a SQL job that does repartitioning a Kafka SQL table into another Kafka SQL table. as example input/output tables have exactly the same SQL schema (see below) and data the only difference is that the new kafka stream need to be repartition using a simple project like item_id (input stream is partitioned by user_id) is there a way to do this via SQL only ? without using org.apache.flink.streaming.connectors.kafka.partitioner.FlinkKafkaPartitioner
In other words how can we express the stream key (keyedBy) via the SQL layer ? For instance in Hive they expose a system column called __key or __partition that can be used to do this via SQL layer (see https://github.com/apache/hive/tree/master/kafka-handler#table-definitions) CREATE TABLE input_kafkaTable ( user_id BIGINT, item_id BIGINT, category_id BIGINT, behavior STRING, ts TIMESTAMP(3) ) WITH ( 'connector' = 'kafka', 'topic' = 'user_behavior_partition_by_uid', 'properties.bootstrap.servers' = 'localhost:9092', ) CREATE TABLE output_kafkaTable ( user_id BIGINT, item_id BIGINT, category_id BIGINT, behavior STRING, ts TIMESTAMP(3) ) WITH ( 'connector' = 'kafka', 'topic' = 'user_behavior_partition_by_iid', 'properties.bootstrap.servers' = 'localhost:9092', ) -- B-Slim _______/\/\/\_______/\/\/\_______/\/\/\_______/\/\/\_______/\/\/\_______
