Hi Constantinos, I agree with David that it is not easily possible to remove a partition while a Flink job is running. Imagine the following scenario:
Your Flink job initially works on 2 partitions belonging to two different topics and you have checkpointing enabled to guarantee exactly-once delivery. It implies that on every checkpoint the offsets of the Kafka topic are stored in a Flink checkpoint to recover from them in case of a failure. Now you trigger the removal of one of the topics and the discovery detects that one of the partitions was removed. If the pipeline now fails before the next checkpoint was taken Flink will try to recover from the previous checkpoint which is invalid by now because the partition is not available anymore. Only if you do not care about loosing data it is possible to simply ignore the removed partition. Best, Fabian
