### AWS

* Add AWS cloud provider option to use an assumed IAM role. For example, this allows running Controller Manager in a account separate from the worker nodes, but still allows all resources created to interact with the workers. ELBs created would be in the same account as the worker nodes for instance.([#59668](https://github.com/kubernetes/kubernetes/pull/59668), [@brycecarman](https://github.com/brycecarman))

*   Fixed iptables FORWARD policy for Docker 1.13 in kubernetes-worker charm ([#54796](https://github.com/kubernetes/kubernetes/pull/54796),[ @Cynerva](https://github.com/Cynerva))

- The DownwardAPIHugePages feature is beta.  Users may use the feature if all workers in their cluster are min 1.20 version.  The feature will be enabled by default in all installations in 1.22. ([#99610](https://github.com/kubernetes/kubernetes/pull/99610), [@derekwaynecarr](https://github.com/derekwaynecarr)) [SIG Node]

- Flag `--concurrent-node-syncs` has been added to cloud node controller which defines how many workers in parallel will be initialising and synchronising nodes. ([#113104](https://github.com/kubernetes/kubernetes/pull/113104), [@pawbana](https://github.com/pawbana)) [SIG API Machinery, Cloud Provider and Scalability]

      - `kubelet_pod_worker_latency_microseconds` -> `kubelet_pod_worker_duration_seconds`
      - `kubelet_pod_start_latency_microseconds` -> `kubelet_pod_start_duration_seconds`
      - `kubelet_cgroup_manager_latency_microseconds` -> `kubelet_cgroup_manager_duration_seconds`
      - `kubelet_pod_worker_start_latency_microseconds` -> `kubelet_pod_worker_start_duration_seconds`

- Added '--concurrent-cron-job-syncs' flag for `kube-controller-manager` to set the number of workers for cron job controller. ([#117550](https://github.com/kubernetes/kubernetes/pull/117550), [@borgerli](https://github.com/borgerli))
- Added '--concurrent-job-syncs' flag for `kube-controller-manager` to set the number of job controller workers. ([#117138](https://github.com/kubernetes/kubernetes/pull/117138), [@tosi3k](https://github.com/tosi3k))

* Fix kubectl describe for pods with controllerRef  ([#45467](https://github.com/kubernetes/kubernetes/pull/45467), [@ddysher](https://github.com/ddysher))
* Skip errors when unregistering juju kubernetes-workers ([#48144](https://github.com/kubernetes/kubernetes/pull/48144), [@ktsakalozos](https://github.com/ktsakalozos))

- "unfinished_work_microseconds" is added to the workqueue metrics; it can be used to detect stuck worker threads. (kube-controller-manager runs many workqueues.) ([#70884](https://github.com/kubernetes/kubernetes/pull/70884), [@lavalamp](https://github.com/lavalamp))

- The `kube-controller-manager` supports `--concurrent-ephemeralvolume-syncs` flag to set the number of ephemeral volume controller workers. ([#102981](https://github.com/kubernetes/kubernetes/pull/102981), [@SataQiu](https://github.com/SataQiu))

- Fixes high CPU usage in kubectl drain ([#95260](https://github.com/kubernetes/kubernetes/pull/95260), [@amandahla](https://github.com/amandahla)) [SIG CLI]
- For vSphere Cloud Provider, If VM of worker node is deleted, the node will also be deleted by node controller ([#92608](https://github.com/kubernetes/kubernetes/pull/92608), [@lubronzhan](https://github.com/lubronzhan)) [SIG Cloud Provider]