You need to enable JavaScript to run this app.
优惠活动
大模型
产品
解决方案
定价
更多

配置带KMS加密卷的Spot Fleet请求时实例启动失败求助

ECS集群Spot实例启动失败:Client.InternalError: Client error on launch 问题排查与修复

尝试为ECS集群启动Spot实例时持续失败,实例启动阶段报错Client.InternalError: Client error on launch。根据官方文档提示,该问题由EBS卷的IAM权限问题导致。已尝试为Fleet角色和实例角色赋予尽可能多的权限(怀疑实例角色是关键,但不确定),但问题仍未解决。以下是最小复现代码(MRE):


最小复现代码(HCL)

IAM 资源

resource "aws_iam_role" "fleet" {
  name = "ecs-fleet"

  assume_role_policy = jsonencode(
    {
      "Version" : "2008-10-17",
      "Statement" : [
        {
          "Sid" : "",
          "Effect" : "Allow",
          "Principal" : {
            "Service" : [
              "spotfleet.amazonaws.com",
              "ec2.amazonaws.com"
            ]
          },
          "Action" : "sts:AssumeRole"
        }
      ]
  })
}

resource "aws_iam_role" "ecs" {
  name = "ecs.role"

  assume_role_policy = jsonencode({
    Version = "2012-10-17"
    Statement = [
      {
        Action = "sts:AssumeRole"
        Effect = "Allow"
        Sid    = ""
        Principal = {
          Service = "ec2.amazonaws.com"
        }
      },
    ]
  })
}

resource "aws_iam_instance_profile" "ecs" {
  name = "ecs.profile"
  role = aws_iam_role.ecs.name
}

resource "aws_iam_policy_attachment" "fleet" {
  name       = "ecs-fleet"
  roles      = [aws_iam_role.fleet.name]
  policy_arn = "arn:aws:iam::aws:policy/service-role/AmazonEC2SpotFleetTaggingRole"
}

KMS 密钥资源

resource "aws_kms_key" "default" {
  policy = jsonencode(
    {
      "Version" : "2012-10-17",
      "Id" : "key-default-1",
      "Statement" : [
        {
          "Sid" : "Allow service-linked role use of the customer managed key",
          "Effect" : "Allow",
          "Principal" : {
            "AWS" : [
              aws_iam_role.ecs.arn,
              aws_iam_role.fleet.arn,
            ]
          },
          "Action" : [
            "kms:Encrypt",
            "kms:Decrypt",
            "kms:ReEncrypt*",
            "kms:GenerateDataKey*",
            "kms:DescribeKey"
          ],
          "Resource" : "*"
        },
        {
          "Sid" : "Allow attachment of persistent resources",
          "Effect" : "Allow",
          "Principal" : {
            "AWS" : [
              aws_iam_role.ecs.arn,
              aws_iam_role.fleet.arn
            ]
          },
          "Action": [
              "kms:CreateGrant",
              "kms:ListGrants",
              "kms:RevokeGrant"
          ],
          "Resource" : "*",
          "Condition" : {
            "Bool" : {
              "kms:GrantIsForAWSResource" : "true"
            }
          },
        }
      ]
    }
  )
}

Spot实例相关资源

resource "aws_launch_template" "ecs_c6i" {
  name = "ECS-c6i.xlarge"
  image_id      = "ami-00f05e43221bf6048" # al2022-ami-ecs-hvm-2022.0.20220630-x86_64-ebs
  ebs_optimized = true
  instance_type = "c6i.xlarge"

  iam_instance_profile {
    name = aws_iam_instance_profile.ecs.name
  }

  block_device_mappings {
    device_name = "/dev/xvda"

    ebs {
      kms_key_id = aws_kms_key.default.arn
      volume_size = 30
      delete_on_termination = true
      encrypted = true
      volume_type = "gp3"
    }
  }
}

resource "aws_spot_fleet_request" "main" {
  iam_fleet_role = aws_iam_role.fleet.arn
  allocation_strategy                 = "lowestPrice"
  target_capacity                     = 1
  terminate_instances_with_expiration = false

  launch_template_config {
    launch_template_specification {
      id      = aws_launch_template.ecs_c6i.id
      version = "$Latest"
    }
  }
}

问题根源与修复方案

核心问题

报错的直接原因是Spot Fleet角色缺少操作KMS密钥的权限,同时KMS密钥策略未允许EC2服务关联角色访问密钥——当Spot Fleet创建加密EBS卷时,需要同时具备这两类权限才能成功挂载卷并启动实例。

修复步骤

  1. 为Spot Fleet角色添加KMS权限
    当前Fleet角色仅附加了AmazonEC2SpotFleetTaggingRole策略,该策略不包含KMS操作权限。需创建自定义IAM策略并附加到Fleet角色:

    # 获取当前AWS账号ID
    data "aws_caller_identity" "current" {}
    
    # 创建KMS访问策略
    resource "aws_iam_policy" "fleet_kms" {
      name        = "ecs-fleet-kms-access"
      description = "允许Spot Fleet使用指定KMS密钥进行EBS加密"
      policy = jsonencode({
        Version = "2012-10-17"
        Statement = [
          {
            Effect = "Allow"
            Action = [
              "kms:Encrypt",
              "kms:Decrypt",
              "kms:ReEncrypt*",
              "kms:GenerateDataKey*",
              "kms:DescribeKey"
            ],
            Resource = aws_kms_key.default.arn
          }
        ]
      })
    }
    
    # 将策略附加到Fleet角色
    resource "aws_iam_policy_attachment" "fleet_kms" {
      name       = "ecs-fleet-kms-attachment"
      roles      = [aws_iam_role.fleet.name]
      policy_arn = aws_iam_policy.fleet_kms.arn
    }
    
  2. 更新KMS密钥策略,添加EC2服务关联角色
    EC2服务在创建加密EBS卷时,会使用自身的服务关联角色AWSServiceRoleForEC2操作KMS密钥,需将该角色加入KMS策略的允许列表:
    修改aws_kms_key.default的policy部分,在两个Statement的Principal.AWS数组中添加以下ARN:

    "arn:aws:iam::${data.aws_caller_identity.current.account_id}:role/aws-service-role/ec2.amazonaws.com/AWSServiceRoleForEC2"
    

    修改后的KMS策略Principal部分示例:

    "Principal" : {
      "AWS" : [
        aws_iam_role.ecs.arn,
        aws_iam_role.fleet.arn,
        "arn:aws:iam::${data.aws_caller_identity.current.account_id}:role/aws-service-role/ec2.amazonaws.com/AWSServiceRoleForEC2"
      ]
    }
    
  3. 验证实例角色权限(可选)
    若后续实例需要直接访问KMS加密资源,可给ECS实例角色添加对应的KMS权限,但这并非本次启动报错的必要条件。


内容的提问来源于stack exchange,提问作者RobinFrcd

相关产品推荐
方舟 Agent Plan

超全模态模型 × Harness 升级,最新支持 Deepseek-V4.1-Flash、GLM-5.3 系列、Doubao-Seedream-5.0-pro、Kimi-K3 (部分), 限时 9.9 元起

最近更新时间:2026.08.21 16:27:38