Question

我一直在使用 AWS ECS 做一个项目，并且工作得很好。但是，当我想破坏架构时遇到了问题。我的 aws_ecs_service 和 module.vpc.aws_internet_gateway 出现循环错误。这是它的外观：

运行 terraform destory

时控制台出错

aws_ecs_service.service: Still destroying... [id=arn:aws:ecs:us-east-1:xxxxxxxxxxxx:service/ecs-cluster/ecs-web-service, 5m10s elapsed]
module.vpc.aws_internet_gateway.this[0]: Still destroying... [id=igw-047a69fdbde0baacb, 5m10s elapsed]

我认为问题源于我创建的 IAM 文件，该文件用于创建然后附加到角色的策略。

asg.tf

data "aws_ami" "amazon_linux" {
  most_recent = true

  filter {
    name    = "name"
    values  = ["amzn-ami*amazon-ecs-optimized"]
  }

  filter {
    name    = "architecture"
    values  = ["x86_64"]
  }

  filter {
    name    = "virtualization-type"
    values  = ["hvm"]
  }
  owners  = ["amazon", "self"]
}

###### ECS launch configuration and ASG
## Creates a template of what the instance will need such as
## AMI, keypair, User Data, instance type, etc.
## Need to work on User Data for setting up URL.
resource "aws_launch_configuration" "lc" {
  name                        = "ecs-lc"
  iam_instance_profile        = aws_iam_instance_profile.ecs_service_role.name
  image_id                    = data.aws_ami.amazon_linux.id
  instance_type               = "t2.micro"
  key_name                    = var.key_name
  security_groups             = [aws_security_group.alb-sg.id]
  associate_public_ip_address = true
  user_data                   = <<EOF
#! /bin/bash
sudo apt-get update
sudo echo "ECS_CLUSTER=${var.cluster_name}" >> /etc/ecs/ecs.config
EOF
  lifecycle {
    create_before_destroy = true
  }
}

## Creates the ASG to scale the instances with a Private instance.
resource "aws_autoscaling_group" "asg" {
  name                      = "asg"
  launch_configuration      = aws_launch_configuration.lc.name
  vpc_zone_identifier       = module.vpc.public_subnets
  min_size                  = 3
  desired_capacity          = 3
  max_size                  = 4
  health_check_type         = "ELB"
  health_check_grace_period = 300
  target_group_arns         = [aws_alb_target_group.alb-target-group.arn]
  protect_from_scale_in     = true

  lifecycle {
    create_before_destroy   = true
  }

  tag {
    key                     = "Name"
    value                   = "dude"
    propagate_at_launch     = true
  }
}

alb.tf

### ALB is used for public hosting since is in a private instance for better security
resource "aws_alb" "alb" {
  name               = "Bevy-ALB"
  load_balancer_type = "application"
  internal           = false
  subnets            = module.vpc.public_subnets
  security_groups    = [aws_security_group.alb-sg.id]

  tags = {
    Environment = "bevy"
    CreatedBy   = "dude"
  }
}

resource "aws_alb_target_group" "alb-target-group" {
  name                = "ECS-Target-Group"
  port                = 80
  protocol            = "HTTP"
  target_type         = "instance"
  vpc_id              = data.aws_vpc.main.id

  health_check {
    path                = "/"
    healthy_threshold   = 2
    unhealthy_threshold = 10
    timeout             = 60
    interval            = 300
    matcher             = "200,301,302"
  }
}

resource "aws_alb_listener" "alb-listener" {
  load_balancer_arn   = aws_alb.alb.arn
  port                = "80"
  protocol            = "HTTP"

  default_action {
    type              = "forward"
    target_group_arn  = aws_alb_target_group.alb-target-group.arn
  }
}

我必须为 Terraform 手动删除 ALB 以销毁剩余资源，但即便如此，我还是遇到了另一个障碍，我无法通过 Terraform 或 AWS 的控制台删除我的 AWS ECS 集群。它说它已被删除但在 AWS 中仍然可用。如果有人可以帮助解决我的问题，我将不胜感激。

编辑：最终能够删除 ECS 集群。我不得不手动删除 ALB 目标组和 ASG。让我更新我的问题以显示我的 ALB 和 ASG。我绝对确定它可能是我的 ASG 中的 lifecycle 用途。

Terraform 无法删除 AWS 中的 ECS 集群

0 个答案: