Public ARN of aws-otel-nodejs-amd64 v1.18.1 is too large

[mt-jmccarthy]

---

name: Bug report
about: Create a report to help us improve
title: 'Public ARN of aws-otel-nodejs-amd64 v1.18.1 is too large'
labels: bug
assignees: ''

---

Describe the bug
Using the public ARN of the aws-otel-nodejs v1.18.1 lambda layer results in AWS CDK deployment failure due to the lambda layer being too large.

Steps to reproduce

• Create an application stack in AWS CDK (v2.125.0)
• Create a reference to the public lambda layer using the ARN arn:aws:lambda:ap-northeast-1:901920570463:layer:aws-otel-nodejs-amd64-ver-1-18-1:1
• Attach that lambda layer reference to a defined lambda function.
• Deploy the application using the AWS CDK command (npx aws-cdk deploy appName)

Sample code

// Typescript (4.9.5)
const getOtelCollectorLayer = (scope: Stack): LayerVersion => {
  return LayerVersion.fromLayerVersionArn(
    scope,
    `OtelCollectorLayer`,
    `arn:aws:lambda:${scope.region}:901920570463:layer:aws-otel-nodejs-amd64-ver-1-18-1:1`
  );
};

export const buildParserHandler = (scope: Stack): Function => {

  const otelLayer = getOtelCollectorLayer(scope);
  
  return new Function(scope, `MyLambdaFunction`, {
    runtime: Runtime.NODEJS_18_X,
    handler: "index.handler",
    memorySize: 256,
    timeout: Duration.seconds(60),
    code: Code.fromAsset(path.join("dist/")),
    layers: [otelLayer],
    functionName: `my-lambda-function}`,
  });
  
};

What did you expect to see?
A successful deployment of the application stack.

What did you see instead?
Deployment fails due to an error contianing:
Resource handler returned message: "Layers consume more than the available size of 262144000 bytes

What version of collector/language SDK version did you use?
Nodejs v1.18.1

What language layer did you use?
Node v18.16.0

Additional context

• I tested deployments with layer version v1.12.0 and v1.17.1, and both deployed successfully. Only v1.18.1 gives me the deployment error.
• There is a known issue with AWS CDK where replacing an existing lambda layer with a new version may cause this error. To work around this, I first deployed a stack where there is no lambda layers, then did another deployment with lambda layer v1.18.1. It still resulted in the same error.