v1.2.1

v1.2.1

Updated

• Create one nodegroup per AZ

v1.2.0

v1.2.0

Added

• Ability to use an existing EKS cluster
• metrics-server and descheduler amenities
• ASG tagging for improved cluster-autoscaler performance
• EKS cluster addon configuration including enableNetworkPolicy=true

Updated

• All amenities to latest versions

v1.1.0

v1.1.0

Updated

• Make toggle variable names more generic and consistent with modules for other cloud providers
• Create GPU node group by default scaled to 0 nodes
• external-dns values now only target ingress resources and use a policy of sync
• ingress-nginx values now correctly set externalTrafficPolicy to Local for internet-facing and allow for empty variables
• Only create public and private DNS zones when needed
• Extensive updates to examples and README
• Make helm charts that use pod identities depend on the kubernetes cluster_addons to ensure that the pod identity agents are installed before the helm charts

v1.0.7

v1.0.7

Updated

• KMS key alias name as a prefix with a unique string appended

v1.0.6

v1.0.6

Added

• eks_primary_nodegroup_name, eks_primary_nodegroup_labels, eks_gpu_nodegroup_name, eks_gpu_nodegroup_labels variables
• nvidia_device_plugin helm chart module

Updated

• Default values for eks_gpu_nodegroup_labels and eks_gpu_nodegroup_taints

v1.0.5

v1.0.5

Added

• vpc_endpoint variable with s3 as the default

v1.0.4

v1.0.4

Fixed

• Removed explicit module.eks dependencies from helm charts which should decrease the likelihood that the aws_eks_cluster_auth token expires before helm charts can be installed

Added

• dns_zone_force_destroy, s3_bucket_force_destroy, and ecr_repositories_force_destroy variables

Updated

• Default IRSA role kubernetes_namespace from dr-core to dr-app

v1.0.3

v1.0.3

Fixed

• Issue where helm charts failed to install on first run

v1.0.2

v1.0.2

Fixed

• Race condition when aws-load-balancer-controller helm chart was installed before its pod identity association was created causing it to lack permissions required to create a load balancer

Added

• Configurable EKS nodegroup taints and AMI types

Updated

• Examples now source the module from the terraform registry rather than locally

v1.0.1

v1.0.1

Added

• Variable eks_cluster_endpoint_private_access_cidrs for users to specify additional CIDR blocks allowed to access the EKS private API endpoint

Updated

• Default eks_primary_nodegroup_instance_types to r6a.4xlarge
• Default eks_primary_nodegroup_desired_size to 5