v14.8.0

Added

• Add missing labels for kube-scheduler and controller-manager.

Fixed

• Fix kube-scheduler healthcheck.

v14.7.0

Fixed

• Allow scraping of kube-scheduler and controller-manager metrics.

Changed

• Revert flatcar to 3227.2.1 to prevent filesystem hang bug (see giantswarm/giantswarm#23831).
• Bump kubernetes to 1.24.8.

v14.6.2

Fixed

• [AWS] Fix IAM permission for master nodes when using cross-account KMS for EBS volumes' encryption.

v14.6.1

• Fix journald systemd override for worker nodes

v14.6.0

Changed

• [Azure] Create CiliumLocalRedirectPolicy for Azure AD pod identity app after creating App CRs to avoid race condition.
• Bump to flatcar 3227.2.3.
• Bump kubernetes to 1.24.7.
• Bump etcd to 3.5.5.
• Bump nginx-ingress-controller-app to version 2.19.0.
• [Azure] Bump azuredisk-csi-driver-app to version v1.21.0-gs4.
• Label master nodes with node-role.kubernetes.io/control-plane to comply with kubeadm/CAPI.

Fixed

• Apply override to journald systemd unit to overcome race condition between auditd and falco that made journald to crash.
• Add 'app-operator.giantswarm.io/version: 0.0.0' label to Azure Apps.
• Create cilium egress whitelist rule.

v14.5.0

Changed

• Bump cilium to 0.5.0.
• Make --enable-priority-and-fairness flag always explicit.
• Set CPU and Memory requests for Api server.

v14.4.1

Fixed

• Use 'MemTotal' rather than 'MemFree' to get memory limit for api server.

v14.4.0

Changed

• Bump cilium to 0.4.0.
• Run EBS controller only on masters on AWS management clusters.

v14.3.0

Changed

• Add *-giantswarm-audit-logs*/* to the list of s3 authorized endpoints accessed from the AWS management clusters.

v14.2.0

Changed

• Bump k8s to 1.24.6.