Hello, I hope you are ok. I have one problem due to cpu busy spin loop waste cycles. My use case is the following, I have used the HybridSource in order to implement a bounded source as an initial source which is modelling a business predicate based on some constraint, and if those constraints gets satisfied, then to move to next source of HybridSource, which is an unbounded kafka source. All worked as expected, but I have the following problem: When I implemented the SourceReader#isAvailable() method I have followed the recommended javadoc - in order to eliminate busy waiting loops (see attached screen #1) But never the less, my cpu is always 100% utilized - busy max (see attached screen #2) Have I missed something? Thanks in advance.