In RabbitMQ streams or Kafka, messages are not deleted after being consumed. If you have a consumer application replicated across multiple Kubernetes pods, how can you ensure that each pod picks up a different message? For instance, if you store the offset in a database, all pods will connect to the same database, sharing the same data. This means they will all see the same next message. What strategies can be employed to prevent this?
Asked
Active
Viewed 176 times
