aws_ecs_service.launch_type can be optional

[asans]

With the introduction of capacity providers, launch_type is not required and can be omitted.

See AWS Docs here:

https://docs.aws.amazon.com/AmazonECS/latest/developerguide/service_definition_parameters.html

Primarily, here is the blurb:

• If launch_type is specified, then capacity_provider_strategy must be omitted.
• If capacity_provider_strategy is specified, then launch_type must be omitted.
• If no capacity_provider_strategy or launch_type is specified, the defaultCapacityProviderStrategy for the cluster is used.

The issue is with the last point. If both launch_type and capacity_provider_strategy is not specified in Terraform, then terraform defaults the launch_type to "EC2". This then causes the capacity provider strategy not to be used or defaulted to the cluster strategy.

So the fix should just be to make the launch_type optional per the AWS service parameter definition and should NOT default to "EC2" when not specified, but rather just omitted when not specified.

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or other comments that do not add relevant new information or questions, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Terraform Version

Terraform v0.12.20

Affected Resource(s)

• aws_ecs_service

[tkrafael]

Also it would be good if it was possible to pass a default value in launch_type. I have following issue:
I have a module that allows dynamic capacity provider as well as launch_type variables.
When I try to omit lauch_type, it won't work as I cannot pass null, empty anything other than "EC2", "FARGATE" or "EXTERNAL".
However, if I pass one of those words AND try to add a capacity provider strategy, I get an error saying I cannot set launch_type and capacity provider together.

Eg:

# assume this is a git module
resource aws_ecs_service this {
  launch_type         = var.launch_type
  dynamic capacity_provider_strategy {
    for_each = toset(var.capacity_provider_strategies)
    content {
      capacity_provider = capacity_provider_strategy.value.capacity_provider
      weight = capacity_provider_strategy.value.weight
      base = capacity_provider_strategy.value.base
    }
  }
# other configuration omitted
}

Terragrunt usage:

terraform {
   # assume it is a git repo
    source = "${include.root.inputs.repo_base}/modules/ecs-anywhere/ecs-service"
}

inputs = {
    service_name = "my"
    task_definition = "myt"
    launch_type = "" # this is part of the problem
    scheduling_strategy = "REPLICA"
    capacity_provider_strategies = [{
        capacity_provider = "FARGATE_SPOT"
        weight = 4
        base = 0
    },
    {
        capacity_provider = "FARGATE"
        weight = 1
        base = 1
    }]
}

[DenisBY]

Do you have any update on this? We have more than 50 services in one cluster with the default_capacity_provider set to FARGATE_SPOT and it would be nice to use default cluster value instead of specifying it for every service.

[DenisBY]

I see it's already fixed. Or I'm looking at the wrong place?
https://github.com/hashicorp/terraform-provider-aws/blob/v4.36.0/internal/service/ecs/service.go#L178

UPDATE:

Ok, I got it. It's optional but default value is EC2. But it should be 'not set'.

[tkrafael]

There's another problem. Setting launch type to null causes a validation error