data aws_ami "eks_worker" is failing in ap-east-1

[vranystepan]

I have issues

We are not able to create EKS cluster in ap-east-1 due to issues with AMI names.

I'm submitting a...

• bug report
• feature request
• support request
• kudos, thank you, warm fuzzy

What is the current behavior?

With any ami we are getting

Error: Your query returned no results. Please change your search criteria and try again.

  on .terraform/modules/<redacted>/terraform-aws-modules-terraform-aws-eks-6ea3582/data.tf line 19, in data "aws_ami" "eks_worker":
  19: data "aws_ami" "eks_worker" {

We suppose that EKS module is always trying to find images (before workers_group_defaults_defaults is being merged with var.workers_group_defaults) with following pattern but apparently there are no such images in the HK region.

amazon-eks-node-${var.cluster_version}-${var.worker_ami_name_filter}

I'm not able to receive any results even with CLI:

aws ec2 describe-images \                                                                       
    --filters \
        Name=owner-id,Values=602401143452 \
        Name=virtualization-type,Values=hvm \
        Name=root-device-type,Values=ebs \
        Name=is-public,Values=true \
        Name=name,Values=amazon-eks-node-\* \
    --output text \
    --query 'max_by(Images[], &CreationDate).ImageId' \
--profile=<redacted> \
--region=ap-east-1
None

Environment details

• Affected module version: v5.1.0
• OS: any
• Terraform version: 0.12.7

Any other relevant info

Perhaps there is different owner in HK region? https://github.com/awslabs/amazon-eks-ami/pull/305/files/fd9cf4458dffa1e9c01a3a42fca1cc7690abc504

[max-rocket-internet]

Perhaps there is different owner in HK region?

Looks like it. Strange.

[vranystepan]

Internally we've applied temp. workaround in data.tf:

...
owners = ["602401143452", "800184023465"]
...

but this is not the solution as 800184023465 does not contain images with relevant names :)

[bsakweson]

I seem to be having the same issue in us-east-1. Here is what I am seeing?

Error: Your query returned no results. Please change your search criteria and try again.

  on .terraform/modules/eks.eks/data.tf line 19, in data "aws_ami" "eks_worker":
  19: data "aws_ami" "eks_worker" {

A pick at the code shows this:

data "aws_ami" "eks_worker" {
  filter {
    name   = "name"
    values = ["amazon-eks-node-${var.cluster_version}-${var.worker_ami_name_filter}"]
  }

  most_recent = true

  # Owner ID of AWS EKS team
  owners = ["602401143452"]
}

I did not add any additional filter.

affected version: v6.0.0
Overriding ami_id with ami-08739803f18dcc019 from this url using workers_group_defaults property does not seem to work either.

[bsakweson]

It turns out my cluster version was incorrect. Cluster version should not contain minor release, it should be for example 1.14. See this https://github.com/awslabs/amazon-eks-ami and this for https://docs.aws.amazon.com/eks/latest/userguide/eks-optimized-ami.html for details.

[JayThomason]

@bsakweson I had the same issue today when I tried upgrading to 1.14.6.

Maybe the docs/README could be updated to reflect that only the major version should be specified.

[barryib]

@vranystepan - Does the @bsakweson's comment help you ?

[barryib]

@bsakweson I had the same issue today when I tried upgrading to 1.14.6.
Maybe the docs/README could be updated to reflect that only the major version should be specified.

This should specified in the AWS EKS documentation. You must only provider major version for EKS cluster. AWS handle minor upgrades for you. You can send a PR for that https://github.com/awsdocs/amazon-eks-user-guide

@vranystepan I'm closing this issue. Feel free to re-open it if the provided "solution" doesn't work for you.

[hato221]

@barryib the issue is not related to EKS version. If you check @vranystepan's aws ec2 describe ... command, you will see that he tried to search for any image which has the owner with ID 602401143452. For Hong Kong (ap-east-1) region, you need to change owner ID to 800184023465 otherwise, your search will return empty results.

[barryib]

@hato221 I opened this PR #1038 to address the default owner issue. Can you please, tell me if it solves your issue.

[sysnasri]

I have the same issue but it was about cluster_version variable which I have set it to 1.22 but it worked with 1.20

[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.