Initially, I would like you all to focus on a few questions before comparing the frameworks:
1. Is there any comparison or similarity between Flink and the Kafka?
2. What could be better in Flink over the Kafka?
3. Is it the problem or system requirement to use one over the other?
I’m generally happy with both technologies as well as Spark Streaming. But as Sourabh points out, there are differences to keep in mind.