ECS fargate service created, but the container is not fully ready

[ghost]

This issue was originally opened by @rajeevr2715 as hashicorp/terraform#19323. It was migrated here as a result of the provider split. The original body of the issue is below.

---

Terraform Version

Terraform v0.11.7

Terraform Configuration Files

resource "aws_alb_target_group" "alb_target_group_cq" {
  name     = "ecs-cq-tg"
  port     = 80
  protocol = "HTTP"
  vpc_id   = "${data.aws_vpc.vpc_data.id}"
  target_type = "ip"

  health_check {
    path             = "/sessions/new"
    matcher          = "200"
    interval         = "200"
    timeout          = "59"
  }
}

resource "aws_ecs_service" "ecs_service_cq" {
  name            = "ecs_service_cq"
  task_definition = "${aws_ecs_task_definition.ecs_task_definition_cq.family}:${max("${aws_ecs_task_definition.ecs_task_definition_cq.revision}", "${data.aws_ecs_task_definition.ecs_task_definition_cq.revision}")}"
  desired_count   = 1
  launch_type     = "FARGATE"
  health_check_grace_period_seconds  = 3000
  cluster =       "${aws_ecs_cluster.ecs_cluster_cq.id}"

  network_configuration {
    security_groups    = ["${aws_security_group.vpc_sg.id}"]
    subnets            = ["${aws_subnet.subnet_for_ecs.*.id}"]
    assign_public_ip = true
  }

  load_balancer {
    target_group_arn = "${aws_alb_target_group.alb_target_group_cq.arn}"
    container_name   = "ecs_container_cq"
    container_port   = "9000"
  }
  depends_on = ["aws_alb_listener.ecs_alb_listener_cq"]
}

resource "null_resource" "configure" {
  provisioner "local-exec" {
    command = "python ${var.configure_cmd} ${aws_lb.alb_ecs_cq.dns_name} "
  }
  depends_on = ["aws_ecs_service.ecs_service_cq"]
}

Debug Output

Crash Output

Expected Behavior

depends_on should check, the container is up and running. ie, here sonarqube is running on ec2 fargate. Sonarqube container will take some time to start up delay. If I add this code at the end of resource

provisioner "local-exec" {
command = "sleep 2m"
}
It will work.

Actual Behavior

Null resource with depends_on to aws_ecs_service will get failed with request timeout. Even the health check interval is not actually wait for the container to fully up and running.

Steps to Reproduce

Additional Context

References

[tomelliff]

I think this is a duplicate of #3107 which has an open but stalled PR at #3485

[bflad]

Thanks for the heads up, @tomelliff! I agree.

@rajeevr2715 this appears to be a similar ask to #3107 so to consolidate discussions and efforts, I'm going to close this issue in preference of the earlier one. I would suggest adding a 👍 reaction to the original issue to help prioritization and following along there. I also fixed the labelling on the old issue so hopefully more folks can find it as an enhancement rather than a question. 😄

[ghost]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.

If you feel this issue should be reopened, we encourage creating a new issue linking back to this one for added context. Thanks!