ACS-ECS-RepairImage.yml

FormatVersion: OOS-2019-06-01
Description:
  zh-cn: 镜像修复, 镜像修复是基于镜像检测的结果对镜像进行离线修复, 关于镜像检测 https://help.aliyun.com/zh/ecs/user-guide/overview-18,
         修复流程会首先基于Alibaba cloud linux镜像创建一台修复实例,将您的镜像的快照创建云盘挂载到修复实例上,离线修复数据盘，然后将修复后的数据盘创建新的镜像。
         一些限制 1)当您的镜像的系统类型是如ubuntu 22.04 64位等公共镜像系统类型时,修复后生成的镜像类型会是如ubuntu_64位自定义镜像系统类型。
  en: Image repair, which is an offline repair of an image based on the results of image check, Please refer to this document for image check https://www.alibabacloud.com/help/en/ecs/user-guide/overview-18
      The repair process will first create a repair instance based on the Alibaba cloud Linux image, create a cloud disk snapshot of your image, mount it to the repair instance, repair the data disk offline, and then create a new image of the repaired data disk.
  name-en: ACS-ECS-RepairImage
  name-zh-cn: 修复镜像
  categories:
    - image_manage
Parameters:
  regionId:
    Type: String
    Label:
      en: RegionId
      zh-cn: 地域ID
    AssociationProperty: RegionId
    Default: '{{ ACS::RegionId }}'
  zoneId:
    Type: String
    Label:
      en:  VSwitch Availability Zone
      zh-cn: 交换机可用区
    AssociationProperty: ALIYUN::ECS::ZoneId
    AssociationPropertyMetadata:
      RegionId: regionId
  repairImageId:
    Type: String
    Label:
      en: ImageId that needs to be repaired
      zh-cn: 修复镜像ID
    AssociationProperty: ALIYUN::ECS::Image::ImageId
    AssociationPropertyMetadata:
      RegionId: regionId
      SupportedImageOwnerAlias:
        - self
      OSType: linux
  baseImageId:
    Type: String
    Label:
      en: BaseImage
      zh-cn: 基础镜像
    Description:
      en: (The basic image is used to create a repair instance. You need to select a basic image with the same architecture as the repair target image. For example, the custom image you want to repair is x86_64, please select aliyun_3_X64 image)
      zh-cn: <font color='red'>基础镜像用来创建修复实例, 您需要选择和修复目标镜像同架构的基础镜像, 如您要修复的自定义镜像是x86_64镜像,请选择 aliyun_3_x64 镜像</font>
    AllowedValues:
      - aliyun_3_x64_20G_alibase_20230727.vhd
      - aliyun_3_arm64_20G_alibase_20230731.vhd
  instanceType:
    Label:
      en: InstanceType
      zh-cn: 实例类型
    Type: String
    AssociationProperty: ALIYUN::ECS::Instance::InstanceType
    AssociationPropertyMetadata:
      RegionId: regionId
      ZoneId: zoneId
  securityGroupId:
    Label:
      en: SecurityGroupId
      zh-cn: 安全组
    Type: String
    AssociationProperty: ALIYUN::ECS::SecurityGroup::SecurityGroupId
    AssociationPropertyMetadata:
      RegionId: regionId
  vSwitchId:
    Label:
      en: VSwitchId
      zh-cn: 交换机
    Type: String
    AssociationProperty: ALIYUN::VPC::VSwitch::VSwitchId
    AssociationPropertyMetadata:
      RegionId: regionId
      ZoneId: zoneId
      Filters:
        - SecurityGroupId: securityGroupId
  systemDiskCategory:
    Label:
      en: SystemDiskCategory
      zh-cn: 系统盘的云盘种类
    Type: String
    AssociationProperty: ALIYUN::ECS::Disk::SystemDiskCategory
    AssociationPropertyMetadata:
      RegionId: regionId
      InstanceType: instanceType
    Default: cloud_essd
  repairItems:
    Label:
      en: RepairItems
      zh-cn: 修复项目
    Type: List
    AllowedValues:
      - all
      - assist
      - aegis
      - growpart
      - virtio
      - selinux
      - dhcp
      - grub
      - fstab
      - nvme
      - cloudinit
    Default: assist
Tasks:
  - Name: checkImage
    Action: ACS::CheckFor
    Description:
      en: Check if the image is x86_ 64 or arm64
      zh-cn: 检查镜像是否为x86_64或arm64
    Properties:
      Service: ECS
      API: DescribeImages
      Parameters:
        RegionId: '{{ regionId }}'
        ImageId: '{{ repairImageId }}'
      DesiredValues:
        - x86_64
        - arm64
      PropertySelector: Images.Image[0].Architecture
    Outputs:
      architecture:
        Type: String
        ValueSelector: Images.Image[0].Architecture
      snapshotId:
        Type: String
        ValueSelector: Images.Image[0].DiskDeviceMappings.DiskDeviceMapping[0].SnapshotId
      imageName:
        Type: String
        ValueSelector: Images.Image[0].ImageName
      osType:
        Type: String
        ValueSelector: Images.Image[0].OSType
      platform:
        Type: String
        ValueSelector: Images.Image[0].Platform
      bootMode:
        Type: String
        ValueSelector: Images.Image[0].BootMode
  - Name: runInstances
    Action: ACS::ExecuteAPI
    Description:
      en: Creates one ECS instances
      zh-cn: 创建一个ECS实例
    Properties:
      Service: ECS
      API: RunInstances
      Parameters:
        RegionId: '{{ regionId }}'
        Amount: 1
        ImageId: '{{ baseImageId }}'
        InstanceType: '{{ instanceType }}'
        InstanceName: 'image_repair_please_no_delete'
        SecurityGroupId: '{{ securityGroupId }}'
        VSwitchId: '{{ vSwitchId }}'
        InternetMaxBandwidthIn: 200
        InternetMaxBandwidthOut: 0
        SystemDiskCategory: '{{ systemDiskCategory }}'
        DataDisk:
          - SnapshotId: '{{ checkImage.snapshotId }}'
            DiskName: 'image_repair_{{ checkImage.snapshotId }}'
            Category: '{{ systemDiskCategory }}'
        InstanceChargeType: PostPaid
    Outputs:
      instanceId:
        Type: String
        ValueSelector: InstanceIdSets.InstanceIdSet[0]
  - Name: untilInstanceReady
    Action: ACS::WaitFor
    Description:
      en: Waits for the created instances to be Running status
      zh-cn: 等待创建的实例进入运行中状态
    Properties:
      Service: ECS
      API: DescribeInstances
      Parameters:
        RegionId: '{{ regionId }}'
        InstanceIds:
          - '{{ runInstances.instanceId }}'
      DesiredValues:
        - Running
      PropertySelector: Instances.Instance[0].Status
  - Name: checkForCreatedDiskReady
    Action: ACS::CheckFor
    Description:
      en: Checks the disk status Available
      zh-cn: 检测云盘的状态是否为可用的
    OnError: stopInstances
    Properties:
      Service: ECS
      API: DescribeDisks
      Parameters:
        RegionId: '{{ regionId }}'
        InstanceId: '{{ runInstances.instanceId }}'
      DesiredValues:
        - In_use
      PropertySelector: Disks.Disk[0].Status
    Outputs:
      diskSN:
        Type: String
        ValueSelector: Disks.Disk[0].SerialNumber
      diskId:
        Type: String
        ValueSelector: Disks.Disk[0].DiskId
  - Name: runCommandOpenselinux
    Action: ACS::ECS::RunCommand
    Description:
      en: Run cloud assistant command on ECS instance
      zh-cn: 在实例中运行命令
    OnError: stopInstances
    Properties:
      regionId: '{{ regionId }}'
      commandContent: sed -i s/^SELINUX=.*/SELINUX=permissive/g /etc/selinux/config
      commandType: RunShellScript
      instanceId: '{{ runInstances.instanceId }}'
    Outputs:
      invocationOutput:
        Type: String
        ValueSelector: invocationOutput
  - Name: rebootInstance
    Action: ACS::ECS::RebootInstance
    Description:
      en: Reboot Instance
      zh-cn: 重启实例
    OnError: stopInstances
    Properties:
      regionId: '{{ regionId }}'
      instanceId: '{{ runInstances.instanceId }}'
  - Name: runCommand
    Action: ACS::ECS::InvokeCommand
    Description:
      en: Run cloud assistant command on ECS instance
      zh-cn: 在实例中运行命令
    OnError: stopInstances
    Properties:
      regionId: '{{ regionId }}'
      commandId: 'ACS-ECS-ImageRepair-for-linux.sh'
      parameters: 
        Fn::Join:
          - ''
          - - '{"repairitems":'
            - '"'
            - Fn::Join:
                - ' '
                - '{{repairItems}}'
            - '"'
            - ',"disk_serial":"'
            - '{{checkForCreatedDiskReady.diskSN}}'
            - '"}'
      instanceId: '{{ runInstances.instanceId }}'
    Outputs:
      invocationOutput:
        Type: String
        ValueSelector: invocationOutput
  - Name: waitDiskFlush
    Action: ACS::Sleep
    Description:
      en: Waiting for IO flash disk
      zh-cn: 等待IO落盘
    Properties:
      Duration: 1M
  - Name: createSnapshot
    Action: ACS::ECS::CreateSnapshot
    Description:
      en: Mount a data disk for the created ECS instance
      zh-cn: 为云盘创建一个快照
    OnError: deleteInstance
    Properties:
      regionId: '{{ regionId }}'
      snapshotName: 'image_repair_{{ checkForCreatedDiskReady.diskId }}'
      diskId: '{{ checkForCreatedDiskReady.diskId }}'
      retentionDays: 1
    Outputs:
      snapshotId:
        Type: String
        ValueSelector: snapshotId
  - Name: createImage
    Action: ACS::ExecuteAPI
    Description:
      en: Creates image
      zh-cn: 用快照创建镜像
    OnError: deleteSnapshot
    OnSuccess: deleteInstance
    Properties:
      Service: ECS
      API: CreateImage
      Parameters:
        RegionId: '{{ regionId }}'
        SnapshotId: '{{ createSnapshot.snapshotId }}'
        ImageName: 
          Fn::Join:
            - '_'
            - - 'repaired'
              - '{{ checkImage.imageName }}'
              - Fn::FormatUTCTime:
                  - '{{ACS::CurrentUTCTime}}'
                  - '%Y%m%d%H%M%S'
        DetectionStrategy: Standard
        Platform:
          Fn::Select:
            - '{{ checkImage.platform }}'
            - Aliyun: Aliyun
              Anolis: Anolis
              CentOS: CentOS
              Ubuntu: Ubuntu
              SUSE: SUSE
              Debian: Debian
              OpenSUSE: OpenSUSE
              Red Hat: RedHat
              Kylin: Kylin
              UOS: UOS
              Fedora: Fedora
              CentOS Stream: CentOS Stream
              AlmaLinux: AlmaLinux
              Rocky Linux: Rocky Linux
              Customized Linux: Customized Linux
              Others Linux: Others Linux
        BootMode: '{{ checkImage.bootMode }}'
        Architecture: '{{ checkImage.architecture }}'
    Outputs:
      imageId:
        Type: String
        ValueSelector: ImageId
  - Name: deleteSnapshot
    Action: ACS::ExecuteAPI
    Description:
      en: Deletes the Snapshot
      zh-cn: 释放创建的快照
    OnError: deleteInstance
    Properties:
      Service: ECS
      API: DeleteSnapshot
      Parameters:
        RegionId: '{{ regionId }}'
        SnapshotId: '{{ createSnapshot.snapshotId }}'
        Force: true
  - Name: stopInstances #远程命令执行失败后等待1分钟再释放，否则报错“IncorrectInstanceStatus.Initializing”（ECS工程师：阿里云的实例有一分钟保护机制，刚创建的实例需要在60s之后才能删除。）
    Action: ACS::Sleep
    Description:
      en: Waiting for instance initialization to complete
      zh-cn: 等待保护机制失效
    Properties:
      Duration: 1M
  - Name: deleteInstance
    Action: ACS::ExecuteAPI
    Description:
      en: Deletes the ECS instance
      zh-cn: 释放创建的ECS实例
    Properties:
      Service: ECS
      API: DeleteInstance
      Parameters:
        RegionId: '{{ regionId }}'
        InstanceId: '{{ runInstances.instanceId }}'
        Force: true
Outputs:
  imageId:
    Type: String
    Value: '{{ createImage.imageId }}'
Metadata:
  ALIYUN::OOS::Interface:
    ParameterGroups:
      - Parameters:
          - regionId
          - repairImageId
          - baseImageId
        Label:
          default:
            zh-cn: 选择镜像
            en: Select Image
      - Parameters:
          - zoneId
          - instanceType
          - securityGroupId
          - vSwitchId
          - systemDiskCategory
        Label:
          default:
            zh-cn: 实例规格
            en: ECS Instance Configure
      - Parameters:
          - repairItems
        Label:
          default:
            zh-cn: 发送远程命令
            en: Run Command