Changes and Releases
Updates on Giant Swarm workload cluster releases, apps, UI improvements and documentation changes.
Changes compared to v33.1.4
Components
- cluster-aws from v6.4.3 to v7.2.5
- Flatcar from v4459.2.1 to v4459.2.2
- Kubernetes from v1.33.6 to v1.34.3
- os-tooling from v1.26.2 to v1.26.3
:warning: Breaking Changes
- The following IAM permissions have been removed from the control plane nodes
- autoscaling:SetDesiredCapacity
- autoscaling:TerminateInstanceInAutoScalingGroup
- Removed
global.providerSpecific.reducedInstanceProfileIamPermissionsForWorkers value, as that’s the default behavior now. It cannot be overridden anymore.
Added
- Add
kubernetes.io/cluster/$clusterName: "owned" and sigs.k8s.io/cluster-api-provider-aws/cluster/$clusterName: "owned" tags to the IRSAClaim CR so that resources created by Crossplane contain the expected tags. This also allows to find the S3 buckets that need to be deleted when removing a cluster. - This change will roll the control plane nodes Add
preKubeadmCommand to wait for the API server load balancer DNS to be resolvable before running kubeadm on control plane nodes. This prevents kubeadm from failing when the ELB DNS record hasn’t propagated yet. - This change will roll the nodes Add Crossplane IAM Roles, policies and instance profiles for worker and control plane nodes. Instead of having an IAM Role per node pool, now we’ll use the same for all node pools.
- Add the
priority-classes default app, enabled by default. This app provides standardised PriorityClass resources like giantswarm-critical and giantswarm-high, which should replace the previous inconsistent per-app priority classes. - This change will roll the nodes on Karpenter node pools Attach the
lb Security Group to Karpenter nodes. - This change will roll the nodes on Karpenter node pools Name instance on AWS after the nodepool name.
Changed
- Chart: Update
cluster to v5.1.2. - Chart: Update
cluster to v5.1.1. - Chart: Update
cluster to v5.1.0. - Chart: Update
cluster to v5.0.0. - Reduce redundant parts of JSON schema for Karpenter vs. MachinePool types of node pools
- Adjust node max pods based on the
nodeCidrMaskSize
Fixed
- Fix Karpenter schema definition: changed from
app schema to helmRelease schema to correctly reflect that Karpenter is deployed as a HelmRelease resource. This fixes incorrect field definitions in extraConfigs (capitalized enum values ConfigMap/Secret and optional field instead of priority). - Fix Karpenter NodePool subnet filtering: when users define custom
subnetTags, the default giantswarm.io/role: "nodes" filter is no longer applied, allowing full control over subnet selection. The cluster ownership tag (sigs.k8s.io/cluster-api-provider-aws/cluster/<cluster-name>: owned) is still enforced for security. - Fix Karpenter HelmRelease: add missing
valuesFrom parent field for extraConfigs, enabling customers to use custom ConfigMaps and Secrets for Karpenter configuration. - Ensure
AWSCluster.spec.network.subnets.tags is not rendered as null - Add missing documentation for node pools (health checks were not listed)
- Ensure defaulting
maxHealthyPercentage since Helm does not use the default from the schema
Removed
- Remove
RolePolicyAttachment crossplane custom resources as they are not needed when using Role and RolePolicy.
Apps
- cert-exporter from v2.9.14 to v2.9.15
- cilium from v1.3.2 to v1.3.4
- cloud-provider-aws from v1.33.2-1 to v2.0.0
- cluster-autoscaler from v1.33.1-2 to v1.34.1-1
- coredns from v1.28.3 to v1.29.1
- etcd-k8s-res-count-exporter from v1.10.11 to v1.10.12
- external-dns from v3.2.0 to v3.4.0
- k8s-audit-metrics from v0.10.10 to v0.10.11
- network-policies from v0.1.1 to v0.1.3
- node-exporter from v1.20.9 to v1.20.10
- Added node-problem-detector v0.5.2
- observability-bundle from v2.3.2 to v2.5.0
- Added priority-classes v0.3.0
- security-bundle from v1.15.0 to v1.16.1
Changed
Changed
Changed
- Chart: Update to upstream v1.34.0.
Changed
- Chart: Update to upstream v1.34.1.
Changed
- Update
coredns image to 1.14.1. - Update
coredns image to 1.14.0.
Changed
Changed
- Sync to upstream helm chart 1.20.0.
- Add option to set annotationPrefix.
- Fixed the missing schema for .provider.webhook.serviceMonitor configs.
- Fixed incorrect indentation of selector labels under spec.template.spec.topologySpreadConstraints when topologySpreadConstraints is set.
- Use kubectl-apply-job when installing CRDs.
- Upgrade external-dns to v0.20.0.
- Update DNSEndpoints CRD.
- Sync to upstream helm chart
1.19.0.- Grant
discovery.k8s.io/endpointslices permission only when using service source. - Update RBAC for
Service source to support EndpointSlices. - Allow extraArgs to also be a map enabling overrides of individual values.
- Set defaults for
automountServiceAccountToken and serviceAccount.automountServiceAccountToken to true in Helm chart values. - Correctly handle
txtPrefix and txtSuffix arguments when both are provided. - Add ability to generate schema with
helm plugin schema. - Regenerate JSON schema with `helm-values-schema-json’ plugin.
- Added ability to configure
imagePullSecrets via helm global value. - Added options to configure
labelFilter and managedRecordTypes via dedicated helm values. - Allow templating
serviceaccount.annotations keys and values, by rendering them using the tpl built-in function. - Added support for
extraContainers argument. - Added support for setting
excludeDomains argument. - Added support for setting
dnsConfig. - Added support for webhook providers.
- Restrict managed record types to A and CNAME.
Changed
Added
Fixed
Removed
- Repository: Remove integration tests.
node-problem-detector v0.5.2
Changed
- Build: Switch to pushing to
default instead of playground catalog as this app will be fully supported in production
Added
- Add KSM metrics
kube_servicemonitor_info and kube_podmonitor_info for ServiceMonitor and PodMonitor resources - Add KSM metrics
kube_podlog_info for PodLog resource
Changed
- Upgrade
kube-prometheus-stack-app to 19.0.0 - Update alloy-app to 0.16.0
Fixed
- Fixed KSM metrics for endpoints
priority-classes v0.3.0
Changed
- Label now uses chart version instead of app version.
Removed
- Removed appVersion (only version is used now).
Changed
- Add missing dependency to all apps.
- Allow to set multiple dependencies on the depends-on annotation.
- Rename
edgedb to gel. - Update
cloudnative-pg (app) to v0.0.12. - Update
gel (app) to v1.0.1.
Changed
- Update Chart.yaml metadata
Changed
- Update chart metadata to add audience and upstream chart version annotations.
Changed
- Update to upstream Helm Chart
v0.1.95 and Zot v2.11.13.
Allow volume expansion in GP3
Changes compared to v33.1.3
Apps
- aws-ebs-csi-driver from v3.3.0 to v3.4.1
Fixed
- Allow volume expansion by default on gp3
- Correct boolean for volume expansion
Changes compared to v33.1.1
Components
- cluster-cloud-director from v2.4.0 to v3.1.2
- Flatcar from v4459.2.1 to v4459.2.2
- Kubernetes from v1.33.6 to v1.34.3
- os-tooling from v1.26.2 to v1.26.3
Added
- Added
fix-dns-nic-allocation.sh Ignition script to attach DNS servers to correct network interfaces. - Add the
priority-classes default app, enabled by default. This app provides standardised PriorityClass resources like giantswarm-critical and giantswarm-high, which should replace the previous inconsistent per-app priority classes. - Add
"helm.sh/resource-policy": keep annotation to VCDCluster CR so that it doesn’t get removed by Helm when uninstalling this chart. The CAPI controllers will take care of removing it, following the expected deletion order.
Changed
- Fix a race condition when populating
/run/metadata/coreos. - Fix race condition in
ntpd unit. - Chart: Update
cluster to v5.1.2. - Chart: Update
cluster to v5.1.1. - Chart: Update
cluster to v5.1.0. - Chart: Update
cluster to v5.0.0.
Apps
- cert-exporter from v2.9.14 to v2.9.15
- cilium from v1.3.2 to v1.3.4
- coredns from v1.28.3 to v1.29.1
- etcd-k8s-res-count-exporter from v1.10.11 to v1.10.12
- network-policies from v0.1.1 to v0.1.3
- node-exporter from v1.20.9 to v1.20.10
- observability-bundle from v2.3.2 to v2.5.0
- Added priority-classes v0.3.0
- security-bundle from v1.15.0 to v1.16.1
Changed
Changed
Changed
- Update
coredns image to 1.14.1. - Update
coredns image to 1.14.0.
Changed
Added
Fixed
Removed
- Repository: Remove integration tests.
Added
- Add KSM metrics
kube_servicemonitor_info and kube_podmonitor_info for ServiceMonitor and PodMonitor resources - Add KSM metrics
kube_podlog_info for PodLog resource
Changed
- Upgrade
kube-prometheus-stack-app to 19.0.0 - Update alloy-app to 0.16.0
Fixed
- Fixed KSM metrics for endpoints
priority-classes v0.3.0
Changed
- Label now uses chart version instead of app version.
Removed
- Removed appVersion (only version is used now).
Changed
- Add missing dependency to all apps.
- Allow to set multiple dependencies on the depends-on annotation.
- Rename
edgedb to gel. - Update
cloudnative-pg (app) to v0.0.12. - Update
gel (app) to v1.0.1.
Changes compared to v33.1.1
Components
- cluster-vsphere from v3.4.0 to v4.1.2
- Flatcar from v4459.2.1 to v4459.2.2
- Kubernetes from v1.33.6 to v1.34.3
- os-tooling from v1.26.2 to v1.26.3
Added
- Add the
priority-classes default app, enabled by default. This app provides standardised PriorityClass resources like giantswarm-critical and giantswarm-high, which should replace the previous inconsistent per-app priority classes. - Add
"helm.sh/resource-policy": keep annotation to VSphereCluster CR so that it doesn’t get removed by Helm when uninstalling this chart. The CAPI controllers will take care of removing it, following the expected deletion order. - Add
"helm.sh/resource-policy": keep annotation to the provider secret. This is to ensure that it isn’t removed by Helm, thus leading to a race condition when deleting the cluster as the vSphere cleaner needs it to clean up resources in vSphere.
Changed
- Chart: Update
cluster to v5.1.2. - Chart: Update
cluster to v5.1.1. - Chart: Update
cluster to v5.1.0. - Chart: Update
cluster to v5.0.0. - Chart: Update
cluster to v4.6.0. - Chart: Update
cluster to v4.5.1. - Chart: Update
cluster to v4.5.0.
Apps
- cert-exporter from v2.9.14 to v2.9.15
- cilium from v1.3.2 to v1.3.4
- cloud-provider-vsphere from v2.0.1 to v2.2.0
- coredns from v1.28.3 to v1.29.1
- etcd-k8s-res-count-exporter from v1.10.11 to v1.10.12
- network-policies from v0.1.1 to v0.1.3
- node-exporter from v1.20.9 to v1.20.10
- observability-bundle from v2.3.2 to v2.5.0
- Added priority-classes v0.3.0
- security-bundle from v1.15.0 to v1.16.1
- vsphere-csi-driver from v3.4.2 to v3.4.3
Changed
Changed
Added
- Add kamaji.enabled value. If set to true, a deployment instead of the dameonset will be used for CPI controller components.
Changed
- Update to upstream
1.34.0.
Changed
- Update
coredns image to 1.14.1. - Update
coredns image to 1.14.0.
Changed
Added
Fixed
Removed
- Repository: Remove integration tests.
Added
- Add KSM metrics
kube_servicemonitor_info and kube_podmonitor_info for ServiceMonitor and PodMonitor resources - Add KSM metrics
kube_podlog_info for PodLog resource
Changed
- Upgrade
kube-prometheus-stack-app to 19.0.0 - Update alloy-app to 0.16.0
Fixed
- Fixed KSM metrics for endpoints
priority-classes v0.3.0
Changed
- Label now uses chart version instead of app version.
Removed
- Removed appVersion (only version is used now).
Changed
- Add missing dependency to all apps.
- Allow to set multiple dependencies on the depends-on annotation.
- Rename
edgedb to gel. - Update
cloudnative-pg (app) to v0.0.12. - Update
gel (app) to v1.0.1.
Changed
- Update upstream chart to
v3.3.1 - Make deployment
affinity and tolerations configurable in values.yaml.
Changes compared to v33.1.1
Components
- cluster-azure from v4.4.0 to v5.1.2
- Flatcar from v4459.2.1 to v4459.2.2
- Kubernetes from v1.33.6 to v1.34.3
- os-tooling from v1.26.2 to v1.26.3
Added
- Add the
priority-classes default app, enabled by default. This app provides standardised PriorityClass resources like giantswarm-critical and giantswarm-high, which should replace the previous inconsistent per-app priority classes. - Add
"helm.sh/resource-policy": keep annotation to AzureCluster CR so that it doesn’t get removed by Helm when uninstalling this chart. The CAPI controllers will take care of removing it, following the expected deletion order.
Changed
- Chart: Update
cluster to v5.1.2. - Chart: Update
cluster to v5.1.1. - Chart: Update
cluster to v5.1.0. - Chart: Update
cluster to v5.0.0.
Apps
- azure-cloud-controller-manager from v1.32.7-1 to v2.0.0
- azure-cloud-node-manager from v1.32.7 to v2.0.0
- azuredisk-csi-driver from v1.32.9 to v2.1.0
- azurefile-csi-driver from v1.32.5 to v2.0.0
- cert-exporter from v2.9.14 to v2.9.15
- cilium from v1.3.2 to v1.3.4
- coredns from v1.28.3 to v1.29.1
- etcd-k8s-res-count-exporter from v1.10.11 to v1.10.12
- external-dns from v3.2.0 to v3.4.0
- k8s-audit-metrics from v0.10.10 to v0.10.11
- network-policies from v0.1.1 to v0.1.3
- node-exporter from v1.20.9 to v1.20.10
- observability-bundle from v2.3.2 to v2.5.0
- Added priority-classes v0.3.0
- security-bundle from v1.15.0 to v1.16.1
Changed
- Chart: Update to upstream v1.34.3. (#132)
Changed
- Chart: Update to upstream v1.34.3. (#118)
Changed
- Chart: Update to upstream v1.34.0. (#118)
- Chart: Update to upstream v1.33.7. (#114)
Changed
- Chart: Update to upstream v1.34.2. (#71)
Changed
Changed
Changed
- Update
coredns image to 1.14.1. - Update
coredns image to 1.14.0.
Changed
Changed
- Sync to upstream helm chart 1.20.0.
- Add option to set annotationPrefix.
- Fixed the missing schema for .provider.webhook.serviceMonitor configs.
- Fixed incorrect indentation of selector labels under spec.template.spec.topologySpreadConstraints when topologySpreadConstraints is set.
- Use kubectl-apply-job when installing CRDs.
- Upgrade external-dns to v0.20.0.
- Update DNSEndpoints CRD.
- Sync to upstream helm chart
1.19.0.- Grant
discovery.k8s.io/endpointslices permission only when using service source. - Update RBAC for
Service source to support EndpointSlices. - Allow extraArgs to also be a map enabling overrides of individual values.
- Set defaults for
automountServiceAccountToken and serviceAccount.automountServiceAccountToken to true in Helm chart values. - Correctly handle
txtPrefix and txtSuffix arguments when both are provided. - Add ability to generate schema with
helm plugin schema. - Regenerate JSON schema with `helm-values-schema-json’ plugin.
- Added ability to configure
imagePullSecrets via helm global value. - Added options to configure
labelFilter and managedRecordTypes via dedicated helm values. - Allow templating
serviceaccount.annotations keys and values, by rendering them using the tpl built-in function. - Added support for
extraContainers argument. - Added support for setting
excludeDomains argument. - Added support for setting
dnsConfig. - Added support for webhook providers.
- Restrict managed record types to A and CNAME.
Changed
Added
Fixed
Removed
- Repository: Remove integration tests.
Added
- Add KSM metrics
kube_servicemonitor_info and kube_podmonitor_info for ServiceMonitor and PodMonitor resources - Add KSM metrics
kube_podlog_info for PodLog resource
Changed
- Upgrade
kube-prometheus-stack-app to 19.0.0 - Update alloy-app to 0.16.0
Fixed
- Fixed KSM metrics for endpoints
priority-classes v0.3.0
Changed
- Label now uses chart version instead of app version.
Removed
- Removed appVersion (only version is used now).
Changed
- Add missing dependency to all apps.
- Allow to set multiple dependencies on the depends-on annotation.
- Rename
edgedb to gel. - Update
cloudnative-pg (app) to v0.0.12. - Update
gel (app) to v1.0.1.
