Noor Muhammad Malik

Manually edited the `kube-proxy-config` `ConfigMap` to set the feature flag `BoundServiceAccountTokenVolume: true` and `kube-proxy` is working now.

controller-manager is crashing with the error: ```bash ubuntu@ip-172-31-21-150:~$ docker logs fcea13fb45e7 W0913 12:28:25.537130 1 feature_gate.go:237] Setting GA feature gate SetHostnameAsFQDN=true. It will be removed in a future release. W0913 12:28:25.537211...

Updated `kube-controller-manager` liveness probe settings with port `10257` and scheme `https`.

Discovered that the `kube-proxy` update is not applied until _after_ the control plane provisioning for `v1.22` is finished. This causes a chicken and egg problem since this update is required...

Observed node rollout to be successful. Following `kube-system` components stuck in `CrashLoopBackOff`: 1. `external-dns`, with error: ```bash nmalik@ZALANDO-70804 ~ % kubectl logs external-dns-68c8f9dd76-lwlf9 -nkube-system time="2022-09-14T17:12:31Z" level=info msg="config: {APIServerURL: KubeConfig: RequestTimeout:30s...

Rebased on `dev` to remove the feature flag from `kube-proxy-config`.

e2e testing fails due to `external-dns` and `vertical-pod-autoscaler` issues. Those are being upgraded. Work will resume on this PR once those upgrades are finished. vpa upgrade: https://github.bus.zalan.do/teapot/issues/issues/3328 external-dns upgrade: #5354

Upgraded to `v1.22.14`. Also, restored the `kuberuntu` version for `v1.21` (which got downgraded mistakenly). This new rebase also includes the `vertical-pod-autoscaler` upgrade so expecting to not see the vpa cause...

The following pods in CrashLoopBackoff in the e2e test: 1. `deployment-status-service` (3 pods) - getting `OOMKilled`. 2. `external-dns` (1 pod) - this is expected to be resolved with the upgrade....

Retriggering seems to have fixed the transient issues observed with `deployment-service-status-service` and `kube-janitor`. Only the `external-dns` seems to be in `CrashLoopBackoff`, which is expected to be the case until the...