in simple scenario where I have: 1 kafka broker N topics No topic partitioning
I want , for example, separate topics between consumers.
I load K consumers(streams), each stream to process N/K topics.
- I can use the same consumer group
- I can load each consumer in its own group
Can you confirm, that rebalance in 2-d case will not happen if new consumer added? This is good, because I would like avoid rebalance.
What are hidden disadvantages of both approaches?