My ECS service failing after task starts

Question

I created ECS service task like these

```
{
    "containerDefinitions": [
        {
            "name": "api",
            "image": "414397229292.dkr.ecr.ap-south-1.amazonaws.com/bids365/backend:79cb8aa",
            "cpu": 256,
            "memoryReservation": 512,
            "portMappings": [
                {
                    "containerPort": 3000,
                    "hostPort": 3000,
                    "protocol": "tcp"
                }
            ],
            "essential": true,
            "environment": [
                {
                    "name": "IMPART_ENV",
                    "value": "dev"
                }
            ],
            "mountPoints": [],
            "volumesFrom": [],
            "disableNetworking": false,
            "privileged": false,
            "readonlyRootFilesystem": true,
            "logConfiguration": {
                "logDriver": "awslogs",
                "options": {
                    "awslogs-group": "bids365/backend/dev",
                    "awslogs-region": "ap-south-1",
                    "awslogs-stream-prefix": "dev"
                }
            },
            "healthCheck": {
                "command": [
                    "CMD-SHELL",
                    "curl -f http://localhost:3000/ping || exit 1"
                ],
                "interval": 60,
                "timeout": 2,
                "retries": 3,
                "startPeriod": 60
            }
        }
    ],
    "family": "service-dev",
    "taskRoleArn": "arn:aws:iam::414397229292:role/task-role-backend-dev",
    "executionRoleArn": "arn:aws:iam::414397229292:role/ecs-task-execution-role",
    "networkMode": "awsvpc",
    "volumes": [],
    "placementConstraints": [],
    "requiresCompatibilities": [
        "FARGATE"
    ],
    "cpu": "256",
    "memory": "512"
}

```

my alb is created like these using terraform

`

resource "aws_lb" "backend" {
  name               = "backend-alb"
  internal           = false
  load_balancer_type = "application"
  security_groups    = [module.sg.id]
  subnets            = ["subnet-082af865c1410b0ef", "subnet-0da703055394aa446", "subnet-02d9bc7c78c939446"]

enable_deletion_protection = false //todo - change this when we get more clarity

tags = {
    Environment = "production"
  }
  lifecycle {
    prevent_destroy = false
  }
}

module "sg" {
  source  = "cloudposse/security-group/aws"
  version = "0.1.3"
  vpc_id = data.aws_vpc.bid365-backend.id
  delimiter = ""
  name = "443-ingress-private-egress"
  rules = [
    {
      type        = "egress"
      from_port   = 0
      to_port     = 65535
      protocol    = "TCP"
      # cidr_blocks = [data.aws_vpc.bid365-backend.cidr_block]
      self = true
    },
    {
      type        = "ingress"
      from_port   = 443
      to_port     = 443
      protocol    = "TCP"
      # cidr_blocks = ["0.0.0.0/0"] 
      self = true
    },
    {
      type        = "ingress"
      from_port   = 80
      to_port     = 80
      protocol    = "TCP"
      # cidr_blocks = ["0.0.0.0/0"] 
      self = true
    },
    {
      type        = "ingress"
      from_port   = 3000
      to_port     = 3000
      protocol    = "TCP"
      # cidr_blocks = ["0.0.0.0/0"] 
      self = true
    }
  ]
}

resource "aws_lb_listener" "redirect_non_ssl" {
  load_balancer_arn = aws_lb.backend.arn
  port              = "80"
  protocol          = "HTTP"

default_action {
    type = "redirect"

redirect {
      port        = "443"
      protocol    = "HTTPS"
      status_code = "HTTP_301"
    }
  }
}

# resource "aws_acm_certificate" "cert" {
#   domain_name       = var.app_dns_entry
#   validation_method = "DNS"

#   lifecycle {
#     prevent_destroy = false
#   }
# }

resource "aws_lb_listener" "app" {
  load_balancer_arn = aws_lb.backend.arn
  port              = "443"
  protocol          = "HTTPS"
  ssl_policy        = "ELBSecurityPolicy-2016-08"
  certificate_arn   = "arn:aws:acm:ap-south-1:414397229292:certificate/4290c5e1-4b49-40bf-afb5-bedeefd072c2"

default_action {
    type = "fixed-response"
    fixed_response {
      content_type = "text/plain"
      message_body = "Not Found
"
      status_code  = "404"
    }
  }

lifecycle {
    prevent_destroy = false
  }
}

`

Everything is correct it seems and task and alb all created but it is failing at health check failed in target group. But the container port is mapped correctly and ping endpoint I created for health check also working correctly if I try to access from the container. Please help me I am stuck here for  a long time and tried almost everything

Answer

Hello.  
Is the path used for ALB health checks correct?  
ALB performs health checks on "/" by default.  
https://docs.aws.amazon.com/elasticloadbalancing/latest/application/target-group-health-checks.html

Answer

Have been in this situation and I feel your pain. Here are a few things that I did

---Check 1

Check the security group inbound on the task/service. And outbound on the ALB

---Check 2

I would try SSHing onto the fargate task. Some helpful instructions on how to do this [here](https://aws.amazon.com/blogs/containers/new-using-amazon-ecs-exec-access-your-containers-fargate-ec2/)

Then increase the healthcheck interval and count on the Targetgroup.

Then once you ssh in figure out of the healthcheck is actually working "http://localhost:3000/ping"

----Check 3

If you can't do check2 maybe try log the ping output in the container for some more clues.

Answer

Hows the security group on the ECS configured?

Answer

Hi,

Can you try without this "healthcheck" part in your Task definition
```
"healthCheck": {
    "command": [
        "CMD-SHELL",
        "curl -f http://localhost:3000/ping || exit 1"
    ],
    ...
}
```
If your task is working, it means that it's the problem.
So probably, you have to allow localhost connection in your security group.
Add this:
```
ingress {
    protocol    = "-1"
    cidr_blocks = ["127.0.0.1/32"]
}
```

By the way, you don't need to put `CMD-SHELL` because the target group also checks this path, it's redondant.

My ECS service failing after task starts

resource "aws_acm_certificate" "cert" {

domain_name = var.app_dns_entry

validation_method = "DNS"

lifecycle {

prevent_destroy = false

}

}

相关内容