Allow for Node Group placement attributes

[reixd]

what

• Add placement variable in order to specify optional placement parameters for the managed nodes (e.g. pin them to a single AZ).
• When the placement variable is set, the module checks if an availability zone is specified, if so, then only the corresponding subnets are passed to the node_group module.

why

• I want a managed node pool with additional placement restrictions, e.g. for EBS volume management.

Example

placement      = {
      availability_zone = "eu-west-1a"
}

module "eks_node_group" {
     #[....]
    name = "alice-pool-euw1a"
    placement = var.placement
     #[....]
}

references

• https://registry.terraform.io/providers/hashicorp/aws/latest/docs/resources/launch_template#placement
• https://registry.terraform.io/providers/hashicorp/aws/latest/docs/resources/eks_node_group

[Gowiem]

This looks good to me other than one question.

cc @Nuru, since you typically have strong opinions on the EKS modules.

[Gowiem]

/test all

[Gowiem]

/test all

[Gowiem]

@reixd we're getting the following errors from our terratests -- Mind looking into it / giving your thoughts?

--- FAIL: TestExamplesComplete (25.13s)
    apply.go:15: 
        	Error Trace:	apply.go:15
        	            				examples_complete_test.go:82
        	Error:      	Received unexpected error:
        	            	FatalError{Underlying: error while running command: exit status 1; 
        	            	Error: Invalid for_each argument
        	            	
        	            	  on ../../main.tf line 60, in data "aws_subnet" "default":
        	            	  60:   for_each = toset(var.subnet_ids)
        	            	
        	            	The "for_each" value depends on resource attributes that cannot be determined
        	            	until apply, so Terraform cannot predict how many instances will be created.
        	            	To work around this, use the -target argument to first apply only the
        	            	resources that the for_each depends on.
        	            	}
        	Test:       	TestExamplesComplete

[Gowiem]

/test terratest

[reixd]

/test terratest

[Gowiem]

/test all

[Gowiem]

@reixd seems to be the same error:

--- FAIL: TestExamplesComplete (23.54s)
    apply.go:15: 
        	Error Trace:	apply.go:15
        	            				examples_complete_test.go:82
        	Error:      	Received unexpected error:
        	            	FatalError{Underlying: error while running command: exit status 1; 
        	            	Error: Invalid for_each argument
        	            	
        	            	  on ../../main.tf line 60, in data "aws_subnet" "default":
        	            	  60:   for_each = toset(var.subnet_ids)
        	            	
        	            	The "for_each" value depends on resource attributes that cannot be determined
        	            	until apply, so Terraform cannot predict how many instances will be created.
        	            	To work around this, use the -target argument to first apply only the
        	            	resources that the for_each depends on.
        	            	}
        	Test:       	TestExamplesComplete

[reixd]

@Gowiem still having this error. I am able to reproduce this error locally, however do not know how to fix it...
Related to hashicorp/terraform#26878
When using the module to install/configure an EKS cluster live it works ¯_(ツ)_/¯

[Gowiem]

@reixd ah interesting. Haven't seen that bug before. I will surface with the contributor team and figure out if we have a way to run tests against 0.14. I don't believe we do, but we should so we'll move this forward one way or another.

[Gowiem]

/test terratest

[Gowiem]

@reixd when you use it locally, what version of terraform are you using? If this introduces an issue which breaks compatibility with 0.13 then we'll need to assess if we want to move it forward or not.

[Gowiem]

/test terratest

[Gowiem]

@reixd can you relax the required_version for this module to < 0.15 and we'll give this a test run on 0.14?

[reixd]

@reixd when you use it locally, what version of terraform are you using?

I do use the latest 0.14 version, however for another reason: optinal variables.

[reixd]

@reixd can you relax the required_version for this module to < 0.15 and we'll give this a test run on 0.14?

Wilco

[Gowiem]

/test all

[Gowiem]

/test all

[Gowiem]

@reixd our unit tests don't like you:

Run # Some legacy support is on 0.11 branches and we determine the Terraform version based on the target branch name
++ cut -d/ -f1
+ VERSION=master
+ [[ master != \0\.\1\1 ]]
+ TF12=0.12.30
++ terraform-0.13 version --json
++ jq -r .terraform_version
+ TF13=0.13.6
++ terraform-0.14 version --json
++ jq -r .terraform_version
+ TF14=0.14.8
++ head -1
+++ jq -r '.required_core[]'
+++ terraform-config-inspect --json .
++ vert -s '>= 0.14.10, < 0.15.0' 0.12.30 0.13.6 0.14.8
+ VERSION=
+ [[ -n '' ]]
Error: Process completed with exit code 1.

That's because you updated to put the minimum version at 0.14.10, which it doesn't seem we're using yet... Can you update to what it was previously but just bump the maximum version to < 0.15

[reixd]

@reixd our unit tests don't like you:

Well if so, I do not like the tests either xD

[reixd]

Using TF 0.13 and changing the subnet IDs to the private ones:

-  subnet_ids         = module.subnets.public_subnet_ids
+  subnet_ids         = module.subnets.private_subnet_cidrs

I get the following error:

                Error Trace:    apply.go:15
                                                        examples_complete_test.go:82
                Error:          Received unexpected error:
                                FatalError{Underlying: error while running command: exit status 1; 
                                Error: InvalidSubnetID.NotFound: The subnet ID '172.16.0.0/19' does not exist
                                        status code: 400, request id: d5c860fc-ee36-48b2-90c9-cf0908789928
                                
                                  on ../../main.tf line 59, in data "aws_subnet" "default":
                                  59: data "aws_subnet" "default"

Maybe there is a problem with the public subnets?

[mergify]

This pull request is now in conflict. Could you fix it @reixd? 🙏

[mergify]

This pull request is now in conflict. Could you fix it @reixd? 🙏

[bridgecrew]

Bridgecrew has found 0 infrastructure configuration error in this PR ⬇️

[bridgecrew]

Bridgecrew has found 0 infrastructure configuration error in this PR ⬇️

[mergify]

This pull request is now in conflict. Could you fix it @reixd? 🙏