v2.11.0

[richardcase]

Release notes for Cluster API Provider AWS (CAPA) v2.11.0

Documentation

Changelog since v2.10.2

Changes by Kind

Support

• Chore(bump): Bump rosa dependency (🌱 chore(bump): Bump rosa dependency #5919, @damdo)
• Chore: bump image-builder to v0.1.48 (🌱 chore: bump image-builder #5859, @damdo)

API Change

• ROSA: Enable setting min=0 for autoscaling nodepools. (🌱 Allow ROSA NodePool autoscaling MinReplicas to be 0 #5816, @jhjaggars)

Feature

• Add support for configuring cluster IAM permissions via Access Entries (✨ feat: Support Access Entries #5583, @joshfrench)
• Cache architecture for each instance type in memory to avoid repeated AWS API requests (✨ Cache architecture for each instance type in memory to avoid unnecessary AWS API requests #5764, @AndiDog)
• Enable IPv6 support for self-managed kubernetes clusters (✨ IPv6 support for self-managed clusters #5603, @tthvo)
• Feat: adding FIPS to ROSAControlPlane (✨ Add the FIPS field to the RosaControlPlane CR. #5873, @tinaafitz)
• Implement nodeadm bootstrapping type (✨ Implement nodeadm bootstrapping type #5700, @faiq)

Bug or Regression

• Enhanced dedicated host logic to require tenancy=host when attempting to use dedicated hosts for instances. (🐛 Fix: Changed dedicated host validation logic to require tenancy=host #5825, @vr4manta)
• Fix constant re-applying of tags if AWSMachine.metadata.annotations is nil (🐛 Fix constant re-applying of tags if AWSMachine.metadata.annotations is nil #5775, @AndiDog)
• Fix: change HostAffinity default 'host'->'default', improve API doc and tests (🐛 fix: change HostAffinity default 'host'->'default' improved API doc and tests #5801, @damdo)
• Fix: use tlsconfig from the manager options for the webhook server (🐛 fix: use tlsconfig from the manager options for the webhook server #5848, @damdo)
• GPU-enabled cluster e2e test is now enabled after confirming ClusterResourceSet handles large ConfigMaps correctly (🐛 fix: Enable GPU-enabled cluster e2e test #5852, @Arpit529Srivastava)
• Print reason for launch template needing update, fix constant instance refreshes for AWSMachinePool (🐛 Print reason for launch template needing update, fix constant instance refreshes for AWSMachinePool by considering SSHKeyName nil and empty string the same #5788, @AndiDog)
• Validate GP3 volume throughput is within the documented range of 125-2000 MiB/s (🐛 Validate GP3 volume throughput is within the documented range of 125-2000 MiB/s #5876, @jhixson74)

Other (Cleanup or Flake)

• Remove the --cloud-provider flag from the apiserver configuration in all templates. The flag defaults to "external" and is no longer needed. Removed in kubernetes v1.33. (🌱 Remove cloud-provider flag usage #5865, @clebs)
• SetupWebhookWithManager methods have been removed from all api packages for all exported types. Additionally, the AWSMachineTemplateWebhook and AWSMachinePoolWebhook types are no longer exported. Callers importing CAPA as a Go library can no longer use these methods and types. (⚠️ Move webhooks out of API packages #5885, @mdbooth)

Uncategorized

• AI policy for the project added. (📖 docs: add AI policy for project #5884, @richardcase)
• Add NestedVirtualization field to CPUOptions in AWSMachineSpec (v1beta2), allowing users to enable or disable nested virtualization on supported instance types (C8i, M8i, R8i). (✨ add NestedVirtualization to CPUOptions in v1beta2 API #5874, @jhjaggars)
• Adding ROSA-HCP LogForwarder AND ROSAMachinePool ImageTypes (✨ ROSA Add logForward config AND ImageTypes #5786, @serngawy)
• Fix AWS SDK request metrics not being captured ([release-2.11] 🐛 Fix AWS SDK request metrics not being captured #5954, @k8s-infra-cherrypick-robot)
• Fix GC not covering all resources by using API response paging ([release-2.11] 🐛 Fix GC not covering all resources by using API response paging #5953, @k8s-infra-cherrypick-robot)
• Fixed nil pointer panic in ROSARoleConfig controller when AWS client initialization failed due to expired credentials or missing region configuration. (🐛 Fix nil pointer panic in ROSARoleConfig when AWS client initialization fails #5945, @andclt)
• Prepare main for the next release (🌱 metadata: prepare main for next release #5785, @chrischdi)
• Templates: bump AWS CCM to v1.32.5 and add variable to control it (🌱 Bump AWS CCM to v1.32.5 and add variable for it in E2Es #5869, @damdo)
• Update GitHub Actions to use SHA hashes ([release-2.11] 🌱 chore: update github workflows to use sha hashes #5950, @k8s-infra-cherrypick-robot)
• Update the image used in cloud build ([release-2.11] 🐛 fix: cloud build image not found #5951, @k8s-infra-cherrypick-robot)

The images for this release are:
registry.k8s.io/cluster-api-aws/cluster-api-aws-controller:v2.11.0

Thanks to all our contributors.

---

This discussion was created from the release v2.11.0.