Creates and EKS Cluster in the given Region.

User can choose:

1. Name of the cluster
2. Number of AZ's to use and therefore desired number of nodes.
3. Minimum number of nodes.
4. Instance type of the nodes.
5. SSH Key to be added to the nodes.

Template uses the new (re:Invent 2019) aws_eks_node_group resource to automate node deployment and cluster join.

Template can be run as follows:

• Locally
• Scalr Next-gen as the remote backend
• Scalr Next-Gen Service Catalog Offering

1. Pull the repo
2. Set variable values in terraform.tvars.(json)
3. Add your AWS access and secret keys to terraform.tfvars(.json), or enter them at the run time prompts. (scalr_aws_access_key, scalr_aws_secret_key)
4. Run terraform init;terraform apply and watch the magic happen.

1. Pull the repo
2. Create a CLI workspace in Scalr Next-Gen and configure the backend to match in main.tf.
3. Create an TF API token in Scalr Next-Gen and add it to ~/.terraformrc.
4. In Scalr Workspace add Terraform variables and values as follows (note that terraform.tfvars(.json) in the template is not used with a remote backend).
   1. region
   2. cluster_name
   3. instance_type
   4. number_of_azs
   5. minimum_nodes
5. Run terraform init;terraform apply and watch the magic happen.

In general follow the example here https://scalr-athena.readthedocs-hosted.com/en/latest/next-gen/service_catalog.html#service-catalog

1. Create Policies (scalr-module.hcl shows the policy bindings that are required)
   1. cloud.locations - Policy to limit the cloud locations (note this can be all locations but the policy must exist)
   2. cloud.instance_types - Restrict the instance types that are allowed. Minimum 4GB of ram.
2. Create a Global Variable name_fmt with REGEX ^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$
3. Create a Global Variable numeric_fmt with REGEX ^[0-9][0-9]*$
4. Fork or clone the Source repo (https://github.com/scalr-eap/aws_eks_cluster)
5. Create the Service Catalog offering pointing to your copy repo
6. Request the offering. It can take 10-15 minutes to deploy the cluster.