In the stream ingestion lab, the architecture involves a producer writing to a Kinesis Data Stream (KDS), and then 2 other KDSs reading from this stream. Why do we need two KDSs? Can we not have just the 2 Kinesis Firehose streams read directly from the producer KDS?
Hello @Ram_Sastry
Sorry for the late reply. To the best of my understanding, the first KDS does some transformation on the raw data from the producer and delivers the transformed data to the 2 other Kinesis Data streams. If we connect these KDSs directly to the producer, we need to repeat the transformation part twice, which is not suitable.
I hope this helps.