bootstrap_env is not working as expected. DNS_CLUSTER_IP env var is not honored by bootstrap.sh script

[kkapoor1987]

Description

While using Managed NodeGroups with create_launch_template: true and passing bootstrap_env it seems like DNS_CLUSTER_IP is not honored in the bootstrap.sh script.

I see values being set in user-data in AWS console

DNS_CLUSTER_IP="192.168.0.10"
SERVICE_IPV4_CIDR="192.168.0.0/24"

bootstrap_env = {
  DNS_CLUSTER_IP = local.DNS_CLUSTER_IP
  SERVICE_IPV4_CIDR = var.ClusterServiceIpv4Cidr
}

Versions

• Terraform: 1.0.8

• Provider(s): 3.63.0

• Module: 17.24.0

Reproduction

Steps to reproduce the behavior:

1. Create a cluster with ClusterService CIDR : 192.168.0.0/24
2. Now check kubelet-config.json on the host or /etc/resolv.conf in pod expected DNS entry should be 192.168.0.10 but will be 10.100.0.10

Code Snippet to Reproduce

Set following in MNG
bootstrap_env = {
DNS_CLUSTER_IP = "192.168.0.10"
}
Create cluster with cluster_service_ipv4_cidr: 192.168.0.0/24

Expected behavior

bootstrap.sh script should set DNS_CLUSTER_IP: 192.168.0.10 in /etc/kubernetes/kubelet/kubelet-config.json

Actual behavior

bootstrap.sh script is setting default DNS_CLUSTER_IP: 10.100.0.10" in kubelet-config.json which is one of the default value in bootstrap.sh
clusterDNS": [
"10.100.0.10"
],

Additional context

TF version 1.0.8
Module version : 17.24.0

[fascenvi]

Maybe I have a similar problem when I tried to change the container_runtime of a custom_ami. The "bootstrap_env" wasn't honored
Just need add the "export" parameter to change the scope of variables to Global and now the bootstrap.sh could read it.

terraform-aws-eks/modules/node_groups/templates/userdata.sh.tpl

Line 23 in d569aa3

${k}="${v}"

22 %{ for k, v in bootstrap_env ~}
23 export ${k}="${v}"
24 %{ endfor ~}

[antonbabenko]

This issue has been resolved in version 18.0.0 🎉

[github-actions]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues. If you have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.