- When the `RetryGenerateName` feature gate is enabled on the kube-apiserver,
  create requests using generateName are retried automatically by the apiserver when the generated name conflicts with an existing resource name, up to a max limit of 7 retries.
  This feature is in alpha. ([#122887](https://github.com/kubernetes/kubernetes/pull/122887), [@jpbetz](https://github.com/jpbetz))

- Fixed a bug in `kube-scheduler` where pending pod preemption caused preemptor pods to be retried more frequently. ([#134245](https://github.com/kubernetes/kubernetes/pull/134245), [@macsko](https://github.com/macsko)) [SIG Scheduling and Testing]

### Bug or Regression

- Client-go: fixes potential data races retrying requests using a custom io.Reader body; with this fix, only requests with no body or with string / []byte / runtime.Object bodies can be retried ([#113933](https://github.com/kubernetes/kubernetes/pull/113933), [@liggitt](https://github.com/liggitt)) [SIG API Machinery]

- Client-go: fixed potential data races retrying requests using a custom `io.Reader` body; with this fix, only requests with no body or with `string` / `[]byte` / `runtime.Object` bodies can be retried ([#113933](https://github.com/kubernetes/kubernetes/pull/113933), [@liggitt](https://github.com/liggitt))

### Other notable changes

* Update to go1.11.5 ([#73326](https://github.com/kubernetes/kubernetes/pull/73326), [@ixdy](https://github.com/ixdy))
* add goroutine to move unschedulable pods to activeq if they are not retried for more than 1 minute ([#72558](https://github.com/kubernetes/kubernetes/pull/72558), [@denkensk](https://github.com/denkensk))

  - PodSecurityPolicy resources will no longer be served from `extensions/v1beta1` in v1.16. Migrate to the `policy/v1beta1` API, available since v1.10. Existing persisted data can be retrieved via the `policy/v1beta1` API.

application/json;as=Table;v=v1alpha1;g=meta.k8s.io. The returned object (if the server supports it) will be of type meta.k8s.io/v1alpha1 with Table, and contain column and row information related to the resource. Each row will contain information about the resource - by default it will be the object metadata, but callers can add the ?includeObject=Object query parameter and receive the full object. In the future kubectl will use this to retrieve the results of `kubectl get`. ([#40848](https://githu...

client can specify a limit to the number of results to return. If more results exist, a token is returned that allows the client to continue the previous list call repeatedly until all results are retrieved.  The resulting list is identical to a list call that does not perform chunking, thanks to capabilities provided by etcd3.  This allows the server to use less memory and CPU when very large lists are returned. This feature is gated as APIListChunking and is not enabled by default. The 1.9 release will...

* Bound cloud allocator to 10 retries with 100 ms delay between retries. ([#61375](https://github.com/kubernetes/kubernetes/pull/61375), [@satyasm](https://github.com/satyasm))

- Updated `IsDNS1123SubdomainWithUnderscore` so that, when it returned an error, it also returned the correct regex information (`dns1123SubdomainFmtWithUnderscore`). ([#132034](https://github.com/kubernetes/kubernetes/pull/132034), [@ChosenFoam](https://github.com/ChosenFoam))