docker_config_json does not work with bootstrap_env vars

[mustafa89]

Description

Passing docker_config_json results in the EKS node not being able to join the cluster. Caused because printf strips the slashes \ in the encoded json.

locals {
docker_config_json = <<JSON
  {
    "bridge": "none",
    "log-driver": "json-file",
    "log-opts": {
      "max-size": "10m",
      "max-file": "10"
    },
    "live-restore": true,
    "max-concurrent-downloads": 10
  }
JSON
}

replace strips the empty spaces.

jsonencode(replace(local.docker_config_json, "/[[:space:]]/", ""))

outputs:

"{\"bridge\":\"none\",\"log-driver\":\"json-file\",\"log-opts\":{\"max-size\":\"10m\",\"max-file\":\"10\"},\"live-restore\":true,\"max-concurrent-downloads\":10}"

When the above is passed to the userdata script.
printf strips the escapes, and export strips the quotes "" resulting in a malformed json object and docker service failing to start.

I've tried all the bash alchemy I know, If there is a way to accomplosh this where i could populate the json object properly in the environment, it would be very nice.

Removing the printf and replacing it with a cat heredoc file insertion could fix this I believe keeping the escaping intact and the export would then inject a properly formated json object in the environment.

⚠️ Note

• Behavior was tested on bash shell. (behaves differently on zsh)
• Tried without jsonencode. tried adding quotes to the json object inside the local as well.
• Not using a custom AMI

Versions

Terraform: 0.13.6
Provider(s): AWS: 3.63.0
Module: 17.23.0

Reproduction

Add the following to the node group definition as well as the local defined in the description

 bootstrap_env = {
           DOCKER_CONFIG_JSON = jsonencode(replace(local.docker_config_json, "/[[:space:]]/", ""))
  }

EKS node won't be able to join the cluster. SSH in to the node and look insde /etc/docker/daemon.json we can see the malformed json.

[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.