Skip to main content

Release Notes v1.71

23KE release notes and upgrade guide​

  • Before upgrade
    • The 23KE configuration chart was unified and moved, so resources need to be annotated to get adopted by the new chart name. To prevent the old charts from deleting resources when they get removed, they need to get suspended first.
    flux suspend hr pre-gardener-configuration
    flux suspend hr gardener-configuration
    kubectl -n flux-system annotate Secret -l helm.toolkit.fluxcd.io/name=pre-gardener-configuration meta.helm.sh/release-name=configuration --overwrite
    kubectl -n flux-system annotate Secret -l helm.toolkit.fluxcd.io/name=gardener-configuration meta.helm.sh/release-name=configuration --overwrite
    kubectl -n garden annotate Secret -l helm.toolkit.fluxcd.io/name=gardener-configuration meta.helm.sh/release-name=certificates --overwrite
    kubectl -n garden annotate certificates.cert.gardener.cloud -l helm.toolkit.fluxcd.io/name=gardener-configuration meta.helm.sh/release-name=certificates --overwrite
    kubectl -n flux-system annotate Certificate -l helm.toolkit.fluxcd.io/name=gardener-configuration meta.helm.sh/release-name=certificates --overwrite
    kubectl -n flux-system annotate Issuer -l helm.toolkit.fluxcd.io/name=gardener-configuration meta.helm.sh/release-name=certificates --overwrite
    If something goes wrong or the charts weren't suspended, other charts might complain about their -base-values Secret missing. To remedy,suspend and then resume the new configuration HelmRelease so it re-generates those Secrets.

Related upstream release notes / changelogs​

Update provider-azure to 1.35.3

[gardener-extension-provider-azure]

🏃 Others​

Update provider-aws to 1.43.2

[gardener-extension-provider-aws]

🏃 Others​

Update provider-gcp to 1.29.3

[gardener-extension-provider-gcp]

🏃 Others​

Update provider-openstack to 1.33.3

[gardener-extension-provider-openstack]

🏃 Others​

Update gardener-controlplane to 1.70.2

[gardener]

⚠️ Breaking Changes​

🐛 Bug Fixes​

  • [USER] A bug has been fixed which could cause kube-proxys from being missing after a Shoot has been woken up from hibernation. (gardener/gardener#7917, @gardener-ci-robot)
  • [OPERATOR] An issue has been fixed that caused traffic from outside of the cluster to Istio-Ingress being blocked. This is only relevant if seed(s) specify additional load balancer annotations via seed.spec.settings.loadBalancerServices.annotations. (gardener/gardener#7911, @gardener-ci-robot)

🏃 Others​

Update cloudprofiles to 0.6.2

What's Changed​

Full Changelog: https://github.com/gardener-community/cloudprofiles/compare/0.6.1...0.6.2

Update provider-alicloud to 1.46.0

[gardener-extension-provider-alicloud]

📖 Documentation​

🏃 Others​

[machine-controller-manager]

⚠️ Breaking Changes​

  • [OPERATOR] Removal of the following flags (and corresponding fields in associated structs): 'machine-creation-timeout' 'machine-drain-timeout', 'machine-pv-detach-timeout', 'machine-health-timeout=10m', 'machine-safety-apiserver-statuscheck-timeout', 'machine-safety-apiserver-statuscheck-period', 'machine-safety-orphan-vms-period', 'machine-max-evict-retries', 'node-conditions', 'bootstrap-token-auth-extra-groups', 'delete-migrated-machine-class'. The MCM no longer accepts these flags since these are options handled by the Machine Controller invoked by platform specific provider launchers. (gardener/machine-controller-manager#769, @elankath)
  • [DEVELOPER] Deletion of 'Driver.GenerateMachineClassForMigration'. Providers need to adapt to this. (gardener/machine-controller-manager#769, @elankath)

✨ New Features​

🐛 Bug Fixes​

📖 Documentation​

🏃 Others​

[terraformer]

🏃 Others​

Docker Images​

gardener-extension-provider-alicloud: eu.gcr.io/gardener-project/gardener/extensions/provider-alicloud:v1.46.0 gardener-extension-admission-alicloud: eu.gcr.io/gardener-project/gardener/extensions/admission-alicloud:v1.46.0

Update shoot-dns-service to 1.34.0

[gardener-extension-shoot-dns-service]

🏃 Others​

Update os-gardenlinux to 0.20.0

[gardener-extension-os-gardenlinux]

🏃 Others​

Update provider-hcloud to 0.6.17

[gardener-extension-provider-hcloud] v0.6.17

Update dashboard to 1.69.1

[dashboard]

⚠️ Breaking Changes​

  • [OPERATOR] The default ingress class annotation under Values.global.dashboard.ingress.annotations['kubernetes.io/ingress.class'] will not be set anymore. Instead, the ingress class name will be set using Values.global.dashboard.ingress.ingressClassName (gardener/dashboard#1499, @petersutter)

🐛 Bug Fixes​

  • [OPERATOR] Fixed an issue where the helm deployment failed with the error annotations.kubernetes.io/ingress.class: Invalid value: "nginx": can not be set when the class field is also set (gardener/dashboard#1499, @petersutter)
Update gardener-controlplane to 1.71.3

[gardener]

🐛 Bug Fixes​

  • [OPERATOR] A bug causing gardenlet to panic when admission-controller is upgraded to v1.71 but gardenlet is still on v1.70. (gardener/gardener#7989, @acumino)
  • [OPERATOR] Several low timeouts (30s) that were introduced in v1.71.0 for several steps are now reverted as in some cases the Network/ControlPlane reconciliation cannot succeed for 30s. (gardener/gardener#8006, @gardener-ci-robot)