Terraform部署ECS服务报错：负载均衡器最多支持5个项

我用单个AWS ALB承载Angular前端（ECS Fargate）和NestJS后端应用，计划配置7个路径模式的监听器规则。执行Terraform apply时触发错误：

updating ECS Service InvalidParameterException: load balancers can have at most 5 items

我清楚ALB监听器规则上限是100，但当前Terraform代码里的ecs_service资源定义了3个静态load_balancer块，加上动态生成的4个，共7个关联项，始终报这个错。

Terraform代码如下：

resource "aws_ecs_service" "ecs_service_def" {
  name            = "backend-service"
  task_definition = "${aws_ecs_task_definition.ecs_task_def.family}"
  desired_count   = 1
  force_new_deployment = true
  health_check_grace_period_seconds  = 300
  launch_type     = "FARGATE"
  cluster         = data.aws_ecs_cluster.this.id
  depends_on      = [
    aws_iam_role_policy.ecs_service_role_policy,
    data.aws_alb_listener.selected443
  ]
  deployment_circuit_breaker {
    enable = true
    rollback = true
  }
  network_configuration {
    security_groups = ["sg-0233333337779"]
    subnets         = var.db_subnet_ids
  }

  load_balancer {
    target_group_arn = aws_alb_target_group.upload.arn
    container_name   = "backend"
    container_port   = "3000"
  }

  load_balancer {
    target_group_arn = aws_alb_target_group.search-logs.arn
    container_name   = "backend"
    container_port   = "3000"
  }

  load_balancer {
    target_group_arn = aws_alb_target_group.log-file-info.arn
    container_name   = "backend"
    container_port   = "3000"
  }


  dynamic "load_balancer" {
    for_each = local.ecs_load_balancers
    content {
      container_name   = load_balancer.value.container_name
      container_port   = load_balancer.value.container_port
      target_group_arn = load_balancer.value.target_group_arn
    }
  }   
}

locals {
  ecs_load_balancers = [
  {
    container_name   = "backend"
    container_port   = "3000"
    target_group_arn = aws_alb_target_group.download.arn
  },
  {
    container_name   = "backend"
    container_port   = "3000"
    target_group_arn = aws_alb_target_group.update-log-status.arn
  },
  {
    container_name   = "backend"
    container_port   = "3000"
    target_group_arn = aws_alb_target_group.log-upload-info.arn
  },
  {
    container_name   = "backend"
    container_port   = "3000"
    target_group_arn = aws_alb_target_group.recent-logs.arn
  }]
}

错误原因 #

你搞混了两个完全独立的AWS限制：

• ALB监听器规则上限是100，这个没问题
• 单个ECS服务最多只能关联5个目标组，这才是触发错误的核心原因，和ALB的规则数量无关

解决方案 #

方案1：合并路径规则到同一目标组（推荐） #

所有目标组都指向同一个容器的3000端口，完全可以把7个路径规则都配置到同一个目标组下：

• 只保留一个目标组关联到ECS服务
• 在ALB监听器上创建7个路径规则，全部转发到这个目标组

修改后的Terraform代码示例：

resource "aws_ecs_service" "ecs_service_def" {
  name            = "backend-service"
  task_definition = aws_ecs_task_definition.ecs_task_def.family
  desired_count   = 1
  force_new_deployment = true
  health_check_grace_period_seconds  = 300
  launch_type     = "FARGATE"
  cluster         = data.aws_ecs_cluster.this.id
  depends_on      = [
    aws_iam_role_policy.ecs_service_role_policy,
    data.aws_alb_listener.selected443
  ]
  deployment_circuit_breaker {
    enable = true
    rollback = true
  }
  network_configuration {
    security_groups = ["sg-0233333337779"]
    subnets         = var.db_subnet_ids
  }

  # 仅保留一个目标组关联
  load_balancer {
    target_group_arn = aws_alb_target_group.backend-main.arn
    container_name   = "backend"
    container_port   = "3000"
  }
}

# 为7个路径创建监听器规则，全部转发到同一个目标组
resource "aws_alb_listener_rule" "backend_paths" {
  for_each = {
    upload            = "/upload/*"
    search-logs       = "/search-logs/*"
    log-file-info     = "/log-file-info/*"
    download          = "/download/*"
    update-log-status = "/update-log-status/*"
    log-upload-info   = "/log-upload-info/*"
    recent-logs       = "/recent-logs/*"
  }

  listener_arn = data.aws_alb_listener.selected443.arn
  # 为每个规则设置唯一优先级，可根据需求调整
  priority     = 10 + index(keys(each.value), each.key) * 10

  action {
    type             = "forward"
    target_group_arn = aws_alb_target_group.backend-main.arn
  }

  condition {
    path_pattern {
      values = [each.value]
    }
  }
}

方案2：拆分ECS服务 #

如果必须保留多个目标组（比如需要不同的健康检查、端口配置），可以把后端服务拆分成多个ECS服务，每个服务关联不超过5个目标组：

• 例如拆成两个服务：backend-service-group1关联3个目标组，backend-service-group2关联4个目标组
• 确保两个服务使用相同的任务定义，ALB规则正确转发到对应目标组

内容的提问来源于stack exchange，提问作者Biplab