### Instrumentation

In v1.9 we focused on improving stability of the components owned by the SIG, including Heapster, Custom Metrics API adapters for Prometheus, and Stackdriver.

### Network

- Fixed a bug where hard evictions due to resource pressure allowed pods
  to use the full termination grace period instead of shutting down instantly. 
  This bug also affected force deleted pods. Both cases now receive a termination grace period of 1 second. ([#124063](https://github.com/kubernetes/kubernetes/pull/124063), [@olyazavr](https://github.com/olyazavr))

- Kubelet did not report `kubelet_volume_stats_*` metrics for Generic Ephemeral Volumes. ([#105569](https://github.com/kubernetes/kubernetes/pull/105569), [@pohly](https://github.com/pohly))
- Kubelet's Node Grace Shutdown will terminate probes when shutting down ([#105215](https://github.com/kubernetes/kubernetes/pull/105215), [@rphillips](https://github.com/rphillips))

- When deleting objects using kubectl with the --force flag, you are no longer required to also specify --grace-period=0. ([#87776](https://github.com/kubernetes/kubernetes/pull/87776), [@brianpursley](https://github.com/brianpursley)) [SIG CLI]

* Pods that are marked for deletion and have exceeded their grace period, but are not yet deleted, no longer count toward the resource quota.([#46542](https://github.com/kubernetes/kubernetes/pull/46542))


#### Dynamic Admission Control

- Respect grace period when updating static pods. ([#106394](https://github.com/kubernetes/kubernetes/pull/106394), [@gjkim42](https://github.com/gjkim42)) [SIG Node and Testing]

* Add error message when trying to use clusterrole with namespace in kubectl ([#36424](https://github.com/kubernetes/kubernetes/pull/36424), [@xilabao](https://github.com/xilabao))

- kube-apiserver: shortening the grace period during a pod deletion no longer moves the `metadata.deletionTimestamp` into the past ([#122646](https://github.com/kubernetes/kubernetes/pull/122646), [@liggitt](https://github.com/liggitt)) [SIG API Machinery]

- The default value for node-monitor-grace-period has been increased to 50s (earlier 40s) (Ref - https://github.com/kubernetes/kubernetes/issues/121793) ([#126287](https://github.com/kubernetes/kubernetes/pull/126287), [@devppratik](https://github.com/devppratik)) [SIG API Machinery, Apps and...

- Deprecated the following kubectl run flags, which are ignored if set: `--cascade`, `--filename`, `--force`, `--grace-period`, `--kustomize`, `--recursive`, `--timeout`, `--wait`. ([#112261](https://github.com/kubernetes/kubernetes/pull/112261), [@brianpursley](https://github.com/brianpursley))