aws_autoscaling_group and aws_launch_config set for destruction 2nd time running

[mahnve]

I have a aws_autoscaling_group and a aws_launch_configuration like so:

resource "aws_launch_configuration" "cyclops_lc" {
  name="dryleaf-cyclops-lc"
  image_id="..."
  instance_type="t2.medium"
  key_name="..."
  security_groups=["...", "..."]
}

resource "aws_autoscaling_group" "cyclops_asg" {
  name="dryleaf-cyclops-asg"
  max_size=6
  min_size=2
  desired_capacity=2
  availability_zones=["us-east-1a", "us-east-1d"]
  health_check_grace_period=300
  force_delete="false"
  default_cooldown=300
  health_check_type="ELB"
  launch_configuration="${aws_launch_configuration.cyclops_lc.name}"
  load_balancers=["${aws_elb.web_elb.name}"]
  vpc_zone_identifier=["...", "..."]
}

running it the first time goes fine. Second time though, it seems to infer changes without the config have changed, which makes it want to recreate the auto scaling group:

$ make tf-apply
mkdir -p build
cd terraform; terraform apply
aws_security_group.nat: Refreshing state... (ID: sg-f3969396)
aws_security_group.db: Refreshing state... (ID: sg-59f6ae3c)
aws_security_group.ssh: Refreshing state... (ID: sg-71c8cc14)
aws_security_group.external_web: Refreshing state... (ID: sg-60d7f005)
aws_security_group.jump: Refreshing state... (ID: sg-51f6ae34)
aws_vpc.dryleaf: Refreshing state... (ID: vpc-129a2c77)
aws_launch_configuration.cyclops_lc: Refreshing state... (ID: dryleaf-cyclops-lc)
aws_internet_gateway.gw: Refreshing state... (ID: igw-7d1fd718)
aws_subnet.db-1d: Refreshing state... (ID: subnet-d1f027a6)
aws_subnet.db-1a: Refreshing state... (ID: subnet-1685744f)
aws_subnet.private-1d: Refreshing state... (ID: subnet-16329661)
aws_subnet.private-1a: Refreshing state... (ID: subnet-1ac50343)
aws_subnet.app-1d: Refreshing state... (ID: subnet-17329660)
aws_subnet.app-1a: Refreshing state... (ID: subnet-1fc50346)
aws_subnet.front-1d: Refreshing state... (ID: subnet-d4f027a3)
aws_subnet.front-1a: Refreshing state... (ID: subnet-09857450)
aws_route_table.public: Refreshing state... (ID: rtb-c55feba0)
aws_instance.wolverine: Refreshing state... (ID: i-e68cc50b)
aws_instance.nat_box: Refreshing state... (ID: i-f13f701c)
aws_route_table_association.app-1a: Refreshing state... (ID: rtbassoc-62009007)
aws_route_table_association.app-1d: Refreshing state... (ID: rtbassoc-61009004)
aws_route_table_association.front-1a: Refreshing state... (ID: rtbassoc-60009005)
aws_route_table_association.front-1d: Refreshing state... (ID: rtbassoc-63009006)
aws_elb.web_elb: Refreshing state... (ID: dryleaf-web)
aws_autoscaling_group.cyclops_asg: Refreshing state... (ID: dryleaf-cyclops-asg)
aws_autoscaling_group.cyclops_asg: Destroying...
aws_security_group.db: Modifying...
  ingress.0.security_groups.#: "0" => "1"
  ingress.0.security_groups.0: "" => "sg-59f6ae3c"
  ingress.0.self:              "1" => ""
aws_security_group.external_web: Modifying...
  ingress.0.cidr_blocks.#: "1" => "0"
  ingress.0.cidr_blocks.0: "0.0.0.0/0" => ""
  ingress.0.from_port:     "80" => "443"
  ingress.0.self:          "0" => ""
  ingress.0.to_port:       "80" => "443"
  ingress.1.cidr_blocks.#: "1" => "0"
  ingress.1.cidr_blocks.0: "0.0.0.0/0" => ""
  ingress.1.from_port:     "443" => "80"
  ingress.1.self:          "0" => ""
  ingress.1.to_port:       "443" => "80"
aws_autoscaling_group.cyclops_asg: Error: ResourceInUse: You cannot delete an AutoScalingGroup while there are instances or pending Spot instance request(s) still in the group.
aws_security_group.db: Modifications complete
aws_security_group.external_web: Modifications complete
aws_launch_configuration.cyclops_lc: Modifying...
  image_id:          "ami-24ac174c" => "ami-24ac174c"
  instance_type:     "t2.medium" => "t2.medium"
  key_name:          "qwaya-default" => "qwaya-default"
  name:              "dryleaf-cyclops-lc" => "dryleaf-cyclops-lc"
  security_groups.#: "1" => "2"
  security_groups.0: "sg-71c8cc14" => "sg-71c8cc14"
  security_groups.1: "" => "sg-60d7f005"
aws_elb.web_elb: Modifying...
  health_check.0.healthy_threshold:   "10" => "3"
  health_check.0.target:              "TCP:80" => "HTTP:80/_monitor"
  health_check.0.timeout:             "5" => "3"
  health_check.0.unhealthy_threshold: "2" => "3"
  listener.0.instance_protocol:       "HTTP" => "http"
  listener.0.lb_protocol:             "HTTP" => "http"
aws_launch_configuration.cyclops_lc: Error: ResourceInUse: Cannot delete launch configuration dryleaf-cyclops-lc because it is attached to AutoScalingGroup dryleaf-cyclops-asg
aws_elb.web_elb: Modifications complete
Error applying plan:

1 error(s) occurred:

* ResourceInUse: You cannot delete an AutoScalingGroup while there are instances or pending Spot instance request(s) still in the group.

Terraform does not automatically rollback in the face of errors.
Instead, your Terraform state file has been partially updated with
any resources that successfully completed. Please address the error
above and apply again to incrementally change your infrastructure.
Makefile:15: recipe for target 'build/terraform_marker_file' failed
make: *** [build/terraform_marker_file] Error 1

[mitchellh]

Looks like a bug. Thanks.

[mitchellh]

Fixed