CloudOps Orchestration Service:ACS-ESS-RollingUpdateByReplaceSystemDiskInScalingGroup

テンプレート名

ACS-ESS-RollingUpdateByReplaceSystemDiskInScalingGroup

今すぐ実行

テンプレートの説明

スケーリンググループ内の ECS インスタンスに対し、システムディスクを交換することでローリングアップデートを実行します。

テンプレートタイプ

自動化

所有者

Alibaba Cloud

入力パラメータ

パラメータ	説明	タイプ	必須	デフォルト値	制限
invokeType	操作タイプ。	String	はい
scalingGroupId	スケーリンググループ ID。	String	はい
enterProcess	実行前に中断するスケーリングプロセス。	List	いいえ	['ScaleIn', 'ScaleOut', 'HealthCheck', 'AlarmNotification', 'ScheduledAction']
exitProcess	実行後に再開するスケーリングプロセス。	List	いいえ	['ScaleIn', 'ScaleOut', 'HealthCheck', 'AlarmNotification', 'ScheduledAction']
lifecycleState	ライフサイクル状態。	List	いいえ	InService
imageId	更新に使用するイメージ ID。	String	いいえ	""
sourceImageId	ロールバックに使用するイメージ ID。	String	いいえ	""
sourceExecutionId	ロールバックに使用する OOS 実行 ID。	String	いいえ	""
scalingConfigurationId	スケーリング設定 ID。	String	いいえ	""
batchPauseOption	一時停止ポリシー。	String	いいえ	Automatic
batchNumber	バッチサイズ。	Number	いいえ	5
maxErrors	バッチ実行ごとに許容される最大エラー数。	Number	いいえ	0
OOSAssumeRole	OOS が引き受ける RAM ロール。	String	いいえ	""

出力パラメータ

パラメータ	説明	タイプ
commandOutputs		List

必要なポリシー

{
    "Version": "1",
    "Statement": [
        {
            "Action": [
                "ess:DescribeScalingConfigurations",
                "ess:DescribeScalingGroups",
                "ess:DescribeScalingInstances",
                "ess:EnterStandby",
                "ess:ExitStandby",
                "ess:ModifyScalingConfiguration",
                "ess:ResumeProcesses",
                "ess:SetInstancesProtection",
                "ess:SuspendProcesses"
            ],
            "Resource": "*",
            "Effect": "Allow"
        },
        {
            "Action": [
                "ecs:DescribeDisks",
                "ecs:DescribeInstances",
                "ecs:DescribeLaunchTemplateVersions",
                "ecs:DescribeLaunchTemplates",
                "ecs:ModifyInstanceAttribute",
                "ecs:ReplaceSystemDisk",
                "ecs:StartInstance",
                "ecs:StopInstance",
                "ecs:TagResources"
            ],
            "Resource": "*",
            "Effect": "Allow"
        },
        {
            "Action": [
                "oos:ListExecutions"
            ],
            "Resource": "*",
            "Effect": "Allow"
        }
    ]
}

関連ドキュメント

ACS-ESS-RollingUpdateByReplaceSystemDiskInScalingGroup

テンプレートの内容

FormatVersion: OOS-2019-06-01
Description:
  en: Rolling update ECS instances that in scaling group by replace system disk
  zh-cn: 通过更换系统盘的方式滚动更新伸缩组内的ECS实例
  name-en: ACS-ESS-RollingUpdateByReplaceSystemDiskInScalingGroup
  name-zh-cn: 通过更换系统盘的方式滚动更新伸缩组内的ECS实例
  categories:
    - elastic_manage
Parameters:
  invokeType:
    Label:
      en: InvokeType
      zh-cn: 操作类型
    Type: String
    AllowedValues:
      - invoke
      - rollback
  scalingGroupId:
    Label:
      en: ScalingGroupId
      zh-cn: 伸缩组ID
    Type: String
  enterProcess:
    Label:
      en: EnterProcess
      zh-cn: 执行开始要暂停的伸缩组流程
    Type: List
    Default:
      - ScaleIn
      - ScaleOut
      - HealthCheck
      - AlarmNotification
      - ScheduledAction
  exitProcess:
    Label:
      en: ExitProcess
      zh-cn: 执行结束要恢复的伸缩组流程
    Type: List
    Default:
      - ScaleIn
      - ScaleOut
      - HealthCheck
      - AlarmNotification
      - ScheduledAction
  lifecycleState:
    Label:
      en: LifecycleState
      zh-cn: 生命周期状态
    Description:
      en: The lifecycle state of the ECS instance to be upgraded
      zh-cn: 要升级的ECS实例所处的生命周期状态
    Type: List
    AllowedValues:
      - InService
      - Protected
    Default: InService
  imageId:
    Label:
      en: InvokeImageId
      zh-cn: 用于更新的镜像
    Type: String
    Default: ''
  sourceImageId:
    Label:
      en: RollbackImageId
      zh-cn: 用于回滚的镜像
    Type: String
    Default: ''
  sourceExecutionId:
    Label:
      en: SourceExecutionId
      zh-cn: 用于回滚的OOS执行ID
    Type: String
    Default: ''
  scalingConfigurationId:
    Label:
      en: ScalingConfigurationId
      zh-cn: 伸缩配置ID
    Type: String
    Default: ''
  batchPauseOption:
    Label:
      en: BatchPauseOption
      zh-cn: 分批模式
    Type: String
    AllowedValues:
      - Automatic
      - FirstBatchPause
      - EveryBatchPause
    Default: Automatic
  batchNumber:
    Label:
      en: BatchNumber
      zh-cn: 分批执行的批次
    Type: Number
    MinValue: 1
    MaxValue: 100
    Default: 5
  maxErrors:
    Label:
      en: MaxErrors
      zh-cn: 分批执行的最大错误次数
    Type: Number
    Default: 0
  OOSAssumeRole:
    Label:
      en: OOSAssumeRole
      zh-cn: OOS扮演的RAM角色
    Type: String
    Default: ''
RamRole: '{{ OOSAssumeRole }}'
Tasks:
  - Name: choiceNextTask
    Action: 'ACS::Choice'
    Description:
      en: Choose next task by invoke type
      zh-cn: 根据执行类型选择任务
    Properties:
      DefaultTask: getInstance
      Choices:
        - When:
            'Fn::Equals':
              - rollback
              - '{{ invokeType }}'
          NextTask: getExecutionInstanceIds
  - Name: getInstance
    Description:
      en: Views the ECS instances
      zh-cn: 获取ECS实例
    Action: 'ACS::ExecuteApi'
    Properties:
      Service: ESS
      API: DescribeScalingInstances
      Parameters:
        ScalingGroupId: '{{ scalingGroupId }}'
        LifecycleStates: '{{ lifecycleState }}'
        CreationTypes:
          - AutoCreated
          - Attached
    Outputs:
      instanceIds:
        Type: List
        ValueSelector: 'ScalingInstances.ScalingInstance | map({"InstanceId": (.InstanceId), "LifecycleState": (.LifecycleState)})[]'
  - Name: whetherTheScalingGroupInstanceIsEmpty
    Action: 'ACS::Choice'
    Description:
      en: Check whether the instances in the services in the scaling group are empty
      zh-cn: 检查伸缩组内服务中的实例是否为空
    Properties:
      DefaultTask: whetherTosuspendProcess
      Choices:
        - When:
            'Fn::Equals':
              - [ ]
              - '{{ getInstance.instanceIds }}'
          NextTask: 'ACS::END'
  - Name: whetherTosuspendProcess
    Action: 'ACS::Choice'
    Description:
      en: Choose next task by process
      zh-cn: 根据流程值判断是否暂停流程
    Properties:
      DefaultTask: suspendProcess
      Choices:
        - When:
            'Fn::Equals':
              - []
              - '{{ enterProcess }}'
          NextTask: whetherToModifyScalingConfiguration
  - Name: suspendProcess
    Action: 'ACS::ExecuteAPI'
    Description:
      en: Suspend process
      zh-cn: 暂停伸缩组流程
    Properties:
      Service: ESS
      API: SuspendProcesses
      Parameters:
        ScalingGroupId: '{{ scalingGroupId }}'
        Process: '{{ enterProcess }}'
  - Name: whetherToModifyScalingConfiguration
    Action: 'ACS::Choice'
    Description:
      en: Choose next task by ScalingConfigurationId
      zh-cn: 根据伸缩配置ID判断是否修改伸缩配置
    Properties:
      DefaultTask: modifyScalingConfiguration
      Choices:
        - When:
            'Fn::Equals':
              - ''
              - '{{ scalingConfigurationId }}'
          NextTask: invokeByReplaceImage
  - Name: modifyScalingConfiguration
    Action: 'ACS::ExecuteAPI'
    Description:
      en: Modify scaling configuration
      zh-cn: 修改伸缩配置
    Properties:
      Service: ESS
      API: ModifyScalingConfiguration
      Parameters:
        ScalingConfigurationId: '{{ scalingConfigurationId }}'
        ImageId: '{{ imageId }}'
  - Name: invokeByReplaceImage
    Action: 'ACS::ESS::ReplaceSystemDisk'
    Description:
      en: Replaces the system disk of ECS instances
      zh-cn: 更换系统盘
    Properties:
      instanceId:
        Fn::Select:
          - 'InstanceId'
          - '{{ ACS::TaskLoopItem }}'
      lifecycleState:
        Fn::Select:
          - 'LifecycleState'
          - '{{ ACS::TaskLoopItem }}'
      imageId: '{{ imageId }}'
      scalingGroupId: '{{ scalingGroupId }}'
    Loop:
      Items: '{{ getInstance.instanceIds }}'
      BatchPauseOption: '{{ batchPauseOption }}'
      MaxErrors: '{{ maxErrors }}'
      Concurrency:
        'Fn::CalculateBatch':
          - '{{ batchNumber }}'
          - '{{ getInstance.instanceIds }}'
      Outputs:
        newSystemDiskIds:
          AggregateType: 'Fn::ListJoin'
          AggregateField: newSystemDiskId
    Outputs:
      newSystemDiskId:
        Type: String
        ValueSelector: newSystemDiskId
  - Name: tagInstance
    Action: ACS::ExecuteApi
    Description:
      en: Creates and attaches tag to ECS instance
      zh-cn: 创建并挂载标签到ECS实例
    Properties:
      Service: ECS
      API: TagResources
      Parameters:
        ResourceIds:
          - Fn::Select:
              - 'InstanceId'
              - '{{ ACS::TaskLoopItem }}'
        ResourceType: Instance
        Tags:
        - Key: 'ess:rollingupdate'
          Value: 'oos:{{ ACS::ExecutionId }}'
    Loop:
      Items: '{{ getInstance.instanceIds }}'
      Mode: Concurrency
      MaxErrors: 0
      Concurrency: 10
  - Name: whetherToResumeProcesses
    Action: 'ACS::Choice'
    Description:
      en: Choose next task by process
      zh-cn: 根据流程值判断是否恢复流程
    Properties:
      DefaultTask: resumeProcesses
      Choices:
        - When:
            'Fn::Equals':
              - []
              - '{{ exitProcess }}'
          NextTask: 'ACS::END'
  - Name: resumeProcesses
    Action: 'ACS::ExecuteAPI'
    OnSuccess: 'ACS::END'
    Description:
      en: Resume process
      zh-cn: 恢复伸缩组流程
    Properties:
      Service: ESS
      API: ResumeProcesses
      Parameters:
        ScalingGroupId: '{{ scalingGroupId }}'
        Process: '{{ exitProcess }}'
  - Name: getExecutionInstanceIds
    Action: 'ACS::ExecuteAPI'
    Description:
      en: Get the instance that has been executed
      zh-cn: 获取已经执行过执行脚本的实例
    Properties:
      Service: OOS
      API: ListExecutions
      Parameters:
        ParentExecutionId: '{{ sourceExecutionId }}'
    Outputs:
      instanceIds:
        Type: List
        ValueSelector: 'Executions[].Parameters.instanceId'
      lifecycleStates:
        Type: List
        ValueSelector: '.Executions | map({"InstanceId": (.Parameters.instanceId), "LifecycleState": (.Parameters.lifecycleState)})[]'
  - Name: getExistInstanceIds
    Description:
      en: Views the ECS instances
      zh-cn: 获取ECS实例
    Action: ACS::SelectTargets
    Properties:
      ResourceType: ALIYUN::ECS::Instance
      Filters:
        - Type: ResourceIds
          ResourceIds: '{{ getExecutionInstanceIds.instanceIds }}'
    Outputs:
      instanceIds:
        Type: List
        ValueSelector: Instances.Instance[].InstanceId
  - Name: getRollbackProcess
    Action: 'ACS::ExecuteAPI'
    Description:
      en: Views resource parameters
      zh-cn: 获取源执行的参数
    Properties:
      Service: OOS
      API: ListExecutions
      Parameters:
        ExecutionId: '{{ sourceExecutionId }}'
    Outputs:
      enterProcess:
        Type: List
        ValueSelector: 'Executions[].Parameters.enterProcess[]'
      exitProcess:
        Type: List
        ValueSelector: 'Executions[].Parameters.exitProcess[]'
  - Name: suspendProcessWithRollBack
    Action: 'ACS::ExecuteAPI'
    Description:
      en: Suspend process
      zh-cn: 暂停伸缩组流程
    Properties:
      Service: ESS
      API: SuspendProcesses
      Parameters:
        ScalingGroupId: '{{ scalingGroupId }}'
        Process: '{{ getRollbackProcess.enterProcess }}'
  - Name: rollbackByReplaceImage
    Action: 'ACS::ESS::ReplaceSystemDisk'
    Description:
      en: Replaces the system disk of ECS instances
      zh-cn: 更换系统盘
    Properties:
      instanceId: '{{ ACS::TaskLoopItem }}'
      lifecycleState:
        Fn::Jq:
          - First
          - '.[] | select(.InstanceId == "{{ ACS::TaskLoopItem }}") | .LifecycleState'
          - '{{ getExecutionInstanceIds.lifecycleStates }}'
      imageId: '{{ sourceImageId }}'
      scalingGroupId: '{{ scalingGroupId }}'
      rollBack: 'true'
    Loop:
      Items: '{{ getExistInstanceIds.instanceIds }}'
      BatchPauseOption: '{{ batchPauseOption }}'
      MaxErrors: '{{ maxErrors }}'
      Concurrency:
        'Fn::CalculateBatch':
          - '{{ batchNumber }}'
          - '{{ getExistInstanceIds.instanceIds }}'
      Outputs:
        newSystemDiskIds:
          AggregateType: 'Fn::ListJoin'
          AggregateField: newSystemDiskId
    Outputs:
      newSystemDiskId:
        Type: String
        ValueSelector: newSystemDiskId
  - Name: whetherToResumeProcessesWithRollback
    Action: 'ACS::Choice'
    Description:
      en: Choose next task by process
      zh-cn: 根据流程值判断是否恢复流程
    Properties:
      DefaultTask: resumeProcessesWithRollback
      Choices:
        - When:
            'Fn::Equals':
              - []
              - '{{ getRollbackProcess.exitProcess }}'
          NextTask: 'ACS::END'
  - Name: resumeProcessesWithRollback
    Action: 'ACS::ExecuteAPI'
    Description:
      en: Resume process
      zh-cn: 恢复伸缩组流程
    Properties:
      Service: ESS
      API: ResumeProcesses
      Parameters:
        ScalingGroupId: '{{ scalingGroupId }}'
        Process: '{{ getRollbackProcess.exitProcess }}'
Outputs:
  commandOutputs:
    Type: List
    Value:
      'Fn::If':
        - 'Fn::Equals':
            - invoke
            - '{{ invokeType }}'
        - '{{ invokeByReplaceImage.newSystemDiskIds }}'
        - '{{ rollbackByReplaceImage.newSystemDiskIds }}'