* The specific behavior of the podGC controller to clean up terminated pods is still governed by the flag, but the podGC's responsibilities have evolved beyond just cleaning up terminated pods.

* Remove headers that are unnecessary for proxy target ([#34076](https://github.com/kubernetes/kubernetes/pull/34076), [@mbohlool](https://github.com/mbohlool))
* Support graceful termination in kube-dns ([#31894](https://github.com/kubernetes/kubernetes/pull/31894), [@MrHohn](https://github.com/MrHohn))

 
## Changes by Kind

### Dependency

- Fix a bug where pod lifecycle hooks could run for their full duration when pods are terminated. ([#136598](https://github.com/kubernetes/kubernetes/pull/136598), [@dgrisonnet](https://github.com/dgrisonnet)) [SIG API Machinery, Auth, Cloud Provider, Node and Scheduling]

### API Change

    SettableFuture<String> resultOfCombiner = SettableFuture.create();
    AsyncCallable<String> combiner =
        tagged(
            "Called my toString",
            () -> {
              // Make this executor terminate after this task so that the test can tell when
              // futureResult has received resultOfCombiner.
              executor.shutdown();
              callableBlocking.await();

    SettableFuture<String> resultOfCombiner = SettableFuture.create();
    AsyncCallable<String> combiner =
        tagged(
            "Called my toString",
            () -> {
              // Make this executor terminate after this task so that the test can tell when
              // futureResult has received resultOfCombiner.
              executor.shutdown();
              callableBlocking.await();