I have a flink 1.16 Job. It has 1.37 GB of full checkpoint data. I have a config:

state.backend.incremental: "true"

But when I check my checkpoint dir I see next: savepoints have weigh of 38GB and checkpoints - 49GB. I had couple of restarts, with restore from savepoint. So I see 4 folders with checkpoints. Can you please suggest, what to do with extra data? Can I detect what I can delete safely? Is there a way for auto clean-up of checkpoint dir by Flink Operator? Thanks in advance