data-on-eks VS terraform-aws-eks-blueprints

git clone https://github.com/awslabs/data-on-eks.git cd data-on-eks/schedulers/terraform/self-managed-airflow

Have you checked out this repo https://github.com/aws-ia/terraform-aws-eks-blueprints

Now that you have the networking part done, you can build configurations for the EKS cluster and its add-ons. You will use the terraform-aws-modules to create the EKS cluster and eks_blueprints module from terraform-aws-eks-blueprintsto configure EKS add-ons.

## (https://github.com/aws-ia/terraform-aws-eks-blueprints) ## ... [other Terraform code] ## Cluster Configuration module "eks" { # ... [other configuration] self_managed_node_groups = { gpu_node_group = { node_group_name = "gpu-node-group" ami_type = "AL2_x86_64_GPU" capacity_type = "ON_DEMAND" instance_types = [ "g4dn.xlarge", "g4dn.2xlarge", ] # ... [other configuration] taints = { dedicated = { key = "nvidia.com/gpu" value = "true" effect = "NO_SCHEDULE" } } # ... [other configuration] } } }

Here is the link: https://github.com/aws-ia/terraform-aws-eks-blueprints/blob/main/examples/karpenter/main.tf

Anyone using eks blueprints or cloudposse's module?

Take a look at the EKS Blueprints for Terraform as a place to start. I know the team is working on their v5 release which should be a solid improvement. https://github.com/aws-ia/terraform-aws-eks-blueprints/milestone/1

Take a look at the EKS Blueprints for Terraform v5 rewrite for more details. However, EKS Blueprints for Terraform (v4 as it is today) is pretty darn good _if_ you just want to manage basic charst like load balancer controller and Karpenter. It provisions IAM Roles and Policies along with Helm charts all in one easy set-up. It's just not something I'd want to touch with more complex use cases and we'll see how the EKS Blueprints team does with the v5 rewrite - their direction looks reasonable, but Terraform just isn't really designed for the problem it's trying to solve there, so it's going to be somewhat clunky one way or another.