Call to function "filebase64sha256" failed: open ../../modules/datadog-lambda-forwarder-dev/forwarder-log.zip: no such file or directory.

[maggie-weber]

Describe the Bug

Any attempts to use this module do not work with the error in the issue title.

╷
│ Error: Error in function call
│ 
│   on .terraform/modules/datadog_lambda_forwarder.forwarder_log_artifact/outputs.tf line 18, in output "base64sha256":
│   18:   value       = module.this.enabled && local.external_curl_filename_effective != "" ? filebase64sha256(local.external_curl_filename_effective) : ""
│     ├────────────────
│     │ while calling filebase64sha256(path)
│     │ local.external_curl_filename_effective is "../../modules/datadog-lambda-forwarder-dev/forwarder-log.zip"
│ 
│ Call to function "filebase64sha256" failed: open
│ ../../modules/datadog-lambda-forwarder-dev/forwarder-log.zip: no such file
│ or directory.

Expected Behavior

The module works.

Steps to Reproduce

Any attempts to use this module resulted in the error text in the "Describe the Bug" section.

Workarounds attempted:

• Downgrading from the most recent version to 1.5.1 (chosen arbitrarily.)
• Changing the variable forwarder_log_artifact_url to a local file:// URL, to the datadog artifact Git repository, and to a direct link to the ZIP artifact file.
• Trying different versions for the variable
• Directly including the curl module as a dependency

Terraform code:

module "datadog_lambda_forwarder" {
    source = "../../modules/datadog-lambda-forwarder-dev"

    dd_api_key_source = {
        resource = "ssm"
        identifier = "/datadog/datadog_api_key"
    }
    dd_forwarder_version = "3.129.0" # Fails either when this value is not provided or when it is
    dd_tags = ["project:${var.environment}-${var.region}-platform-waf"]
    environment = var.environment
    forwarder_log_enabled = true
    name = "Datadog-Lambda-Forwarder-xxxxxx"
    s3_buckets = [ module.waf_log_bucket.bucket_id ]
}

Screenshots

No response

Environment

• OS: ruby:latest container, so Debian.
• Module version: 1.7.0, 1.5.1
• Terraform version: OpenTofu, last updated October 28

We also wrap the build process with Terraspace.

Additional Context

No response

[jnewton-tw]

This is also affecting me. I use Gitlab CI pipeline to do the terraform plan/apply jobs.. the plan job stores the artifacts which include this directory, but the apply still fails. What's the recommended way to approach this? Pre-create these files? I thought the terraform init should do this.

[maggie-weber]

This is also affecting me. I use Gitlab CI pipeline to do the terraform plan/apply jobs.. the plan job stores the artifacts which include this directory, but the apply still fails. What's the recommended way to approach this? Pre-create these files? I thought the terraform init should do this.

I was never able to get this module to work. Instead, my team opted for the more-complicated-but-actually-functional approach of using the datadog_integration_aws Terraform resource.

resource "aws_cloudformation_stack" "datadog_forwarder" {
    name = "datadog-forwarder"
    capabilities = ["CAPABILITY_IAM", "CAPABILITY_NAMED_IAM", "CAPABILITY_AUTO_EXPAND"]
    parameters = {
        DdApiKey = var.datadog_api_key
        DdSite = "datadoghq.com"
        DdTags = "env:${var.environment}"
        FunctionName = "datadog-forwarder"
    }
    template_url = "https://datadog-cloudformation-template.s3.amazonaws.com/aws/forwarder/latest.yaml"
}

It's been long enough that I don't remember all of the details here, but hopefully this should get you started.

[jnewton-tw]

Thank you, i actually got this fixed by making sure it's captured as an artifact by the pipeline.