- Added new annotation `batch.kubernetes.io/cronjob-scheduled-timestamp` to Job objects scheduled from CronJobs. ([#118137](https://github.com/kubernetes/kubernetes/pull/118137), [@helayoty](https://github.com/helayoty))

* Upgrade Stackdriver Logging Agent addon image to 0.6-1.6.0-1 to use Fluentd v1.2. This provides nanoseconds timestamp granularity for logs. ([#70954](https://github.com/kubernetes/kubernetes/pull/70954), [@qingling128](https://github.com/qingling128))

  `terminating` is true when an endpoint is terminating. For pods this is any endpoint with a deletion timestamp. ([#92968](https://github.com/kubernetes/kubernetes/pull/92968), [@andrewsykim](https://github.com/andrewsykim)) [SIG Apps and Network]
- Add dual-stack Services (alpha).  This is a BREAKING CHANGE to an alpha API.

- Upgrade Stackdriver Logging Agent addon image to 0.6-1.6.0-1 to use Fluentd v1.2. This provides nanoseconds timestamp granularity for logs. ([#70954](https://github.com/kubernetes/kubernetes/pull/70954), [@qingling128](https://github.com/qingling128))

  * Fix a bug where container cannot run as root when SecurityContext.RunAsNonRoot is false. ([#47009](https://github.com/kubernetes/kubernetes/pull/47009), [@yujuhong](https://github.com/yujuhong))

  * Fix the Kubelet PLEG update timestamp to better reflect the health of the component when the container runtime request hangs. ([#45496](https://github.com/kubernetes/kubernetes/pull/45496), [@andyxning](https://github.com/andyxning))

- Clean ReplicaSet by revision instead of creation timestamp in deployment controller ([#97407](https://github.com/kubernetes/kubernetes/pull/97407), [@waynepeking348](https://github.com/waynepeking348)) [SIG Apps]

- kube-apiserver now reloads the `--authorization-config` file when it changes. Reloads increment the `apiserver_authorization_config_controller_automatic_reload_last_timestamp_seconds` timestamp metric, with `status="success"` for successful reloads and `status="failed"` for failed reloads. Failed reloads keep using the previously loaded authorization configuration.

- Fixed the bug where Events failed to be created when the referenced object name was not a valid Event name. Now, a UUID is used as the name instead of the referenced object name and the timestamp suffix. ([#129790](https://github.com/kubernetes/kubernetes/pull/129790), [@aojea](https://github.com/aojea))

* Upgrade Stackdriver Logging Agent addon image to 0.6-1.6.0-1 to use Fluentd v1.2. This provides nanoseconds timestamp granularity for logs. ([#70954](https://github.com/kubernetes/kubernetes/pull/70954), [@qingling128](https://github.com/qingling128))

- Fixed bug that kubelet resource metric `container_start_time_seconds` had timestamp equal to container start time. ([#120518](https://github.com/kubernetes/kubernetes/pull/120518), [@saschagrunert](https://github.com/saschagrunert)) [SIG Instrumentation, Node and Testing]