pulumi-eks

pulumi-eks copied to clipboard

What happened?

Have two separate projects, one with the EKS cluster in and one with v2 NodeGroup.

When running pulumi up against the second stack, I get the error:

error: Error: invocation of aws:ssm/getParameter:getParameter returned an error: invoking aws:ssm/getParameter:getParameter: 1 error occurred:
        * Error describing SSM parameter (/aws/service/eks/optimized-ami/undefined/amazon-linux-2/recommended/image_id): ParameterNotFound:

(note the undefined in the SSM parameter path - it should be a version number)

Expected Behavior

The NodeGroup should be deployed.

Steps to reproduce

Code for the EKS cluster stack and the nodegroup stack: https://gist.github.com/pierskarsenbarg/94c004e7756b357f06584b02273bc54f

Output of pulumi about

CLI          
Version      3.69.0
Go Version   go1.20.4
Go Compiler  gc

Plugins
NAME        VERSION
aws         5.41.0
aws         5.10.0
awsx        1.0.2
docker      3.6.1
eks         0.42.7
kubernetes  3.20.2
nodejs      unknown

Host     
OS       darwin
Version  13.4
Arch     x86_64

Current Stack: pierskarsenbarg/node-stack/dev

Found no resources associated with dev

Found no pending operations associated with dev

Backend        
Name           pulumi.com
URL            https://app.pulumi.com/pierskarsenbarg
User           pierskarsenbarg
Organizations  pierskarsenbarg, karsenbarg, team-ce, demo

Additional context

No response

Contributing

Vote on this issue by adding a 👍 reaction. To contribute a fix for this issue, leave a comment (and link to your pull request, if you've opened one already).

Note this also happens in v1 of the EKS SDK as well

Hi @pierskarsenbarg, reviewing the code I see that the missing version is supposed to come from one of two places: the NodeGroupV2's arguments (optional), or else the cluster version. Is this maybe the first time you're up-ing this stack, and there is no cluster yet?

As a workaround, you should be able to specify Version.

@thomas11 the cluster does exist. I believe the fact that it’s not picking up the version from the cluster to be because of #855

I see, @pierskarsenbarg. #855 looks like a plausible cause. Any luck specifying an explicit version as a workaround?

@thomas11 what's the input for that? version looks to be the kubernetes version. Do we then match that up with the corresponding AMI version?

@pierskarsenbarg reading the code, it seems that it is the k8s version that's used in the AMI URL. Unless I'm misunderstanding something - I'm not super familiar with EKS.

    const version = pulumi.output(args.version || core.cluster.version);

    // https://docs.aws.amazon.com/eks/latest/userguide/retrieve-ami-id.html
    let amiId: pulumi.Input<string> | undefined = args.amiId;
    if (!amiId) {
        const amiType = args.amiType ?? args.gpu ? "amazon-linux-2-gpu" : "amazon-linux-2";
        amiId = version.apply((v) => {
            const parameterName = `/aws/service/eks/optimized-ami/${v}/${amiType}/recommended/image_id`;
            //                                                    ^^^^
            return pulumi.output(
                aws.ssm.getParameter({ name: parameterName }, { parent, async: true }),
            ).value;
        });
    }

@thomas11 I don't believe it is the k8s version. Looking at the actual list in the AWS Console, the latest for the amazon-linux-v2 that are in existence are the following:

which was last modified a while ago.

Then there are the arm64 versions:

@pierskarsenbarg - there is a PR open to fix this here. Could someone from Pulumi take a look, please?