All Products
Search
Document Center

Auto Scaling:Automatically unbind secondary ENIs from ECS instances

Last Updated:Dec 28, 2023

This topic describes how to use the lifecycle hook feature of Auto Scaling to put Elastic Compute Service (ECS) instances into the Pending state and then use a CloudOps Orchestration Service (OOS) template to automatically unbind secondary elastic network interfaces (ENIs) that have elastic IP addresses (EIPs) from the ECS instances.

Prerequisites

  • An Alibaba Cloud account is created. To create an Alibaba Cloud account, go to the Sign up to Alibaba Cloud page.

  • A scaling group is created and enabled.

  • An ECS instance of the scaling group is bound with a secondary ENI that has an EIP.

  • A RAM role is created for OOS. The trusted entity type of the RAM role is Alibaba Cloud Service. The trusted service is CloudOps Orchestration Service. The RAM role has the permissions to perform O&M operations that are defined in OOS templates. For more information, see Grant RAM permissions to OOS.

    Note

    In this topic, the OOSServiceRole RAM role is used as an example. You can also use other roles.

Background information

By default, when you delete an ECS instance that is bound with a secondary ENI that has an EIP, the ENI is released but the EIP of the ENI remains. You can use the lifecycle hook feature of Auto Scaling and an OOS template to automatically unbind the secondary ENI from the ECS instance and delete the EIP of the ENI at the same time.

Procedure

In this example, the ACS-ESS-LifeCycleDetachNetworkInterfaceAndDeleteEip public template is used to show how to unbind a secondary ENI from an ECS instance during a scale-in and release the EIP of the ENI at the same time.

Step 1: Grant OOS permissions to the RAM role

You must have the permissions to execute OOS templates. In this example, the ACS-ESS-LifeCycleDetachNetworkInterfaceAndDeleteEip public template is used. The template defines the ECS, Auto Scaling, and EIP resources that are required to perform O&M operations.

  1. Log on to the RAM console.

  2. In the left-side navigation pane, choose Identities > .

  3. Find the RAM role OOSServiceRole and click Add Permissions in the Actions column.

    Grant the required permissions to OOSServiceRole.

  4. In the Add Permissions panel, configure parameters as prompted and click OK.

    The following table describes the parameters that are used in this example. For parameters that are not described in the table, use the default settings.

    Parameter

    Description

    Authorized Scope

    Select Alibaba Cloud Account.

    Select Policy

    Select the following system policies: AliyunECSFullAccess, AliyunESSFullAccess, and AliyunEIPFullAccess.

Step 2: Create a lifecycle hook and trigger a scale-in

  1. Log on to the Auto Scaling console.

  2. In the left-side navigation pane, click Scaling Groups.

  3. In the top navigation bar, select the region where Auto Scaling is activated.

  4. Find a scaling group and use one of the following methods to go to the scaling group details page:

    • Click the ID of the scaling group in the Scaling Group Name/ID column.

    • Click Details in the Actions column.

  5. Create a lifecycle hook.

    1. In the upper part of the scaling group details page, click the Lifecycle Hook tab.

    2. Click Create Lifecycle Hook.

    3. Configure parameters of the lifecycle hook and click OK.

      The following table describes the parameters that are used in this example. For parameters that are not described in the table, use the default settings.

      Parameter

      Description

      Name

      Enter ESSHookForDeleteNicAndEip.

      Scaling Activity

      Select Scale-in Event.

      Timeout Period

      Configure the Timeout Period parameter based on your business requirements. Unit: seconds. In this example, the Timeout Period parameter is set to 300.

      Note

      The timeout period is the period of time during which you can perform custom operations on ECS instances. If the timeout period is shorter than the period of time that is required to perform the custom operations, the operations may fail. We recommend that you estimate the period of time that is required to perform custom operations on ECS instances and configure the Timeout Period parameter based on your business requirements.

      Default Execution Policy

      Select Continue.

      Send Notification When Lifecycle Hook Takes Effect

      Specify a notification method or the action that you want Auto Scaling to perform after the lifecycle hook times out. In this example, perform the following operations:

      • Select OOS Template.

      • Select Public Templates.

      • Select ACS-ESS-LifeCycleDetachNetworkInterfaceAndDeleteEip.

      In the ACS-ESS-LifeCycleDetachNetworkInterfaceAndDeleteEip public template, set OOSAssumeRole to OOSServiceRole. In Step 1: Grant OOS permissions to the RAM role, OOSServiceRole is granted the permissions on the ECS, Auto Scaling, and EIP resources. OOS obtains the preceding permissions after it assumes the RAM role.

  6. Trigger a scale-in.

    In this example, a scale-out is manually triggered by executing a scaling rule. You can also trigger scale-outs by using scheduled or event-triggered tasks.

    Note

    If scaling activities are triggered when you manually execute scaling rules, lifecycle hooks take effect. However, lifecycle hooks do not take effect when you manually add or remove ECS instances to or from a scaling group.

    1. In the upper part of the scaling group details page, click the Scaling Rules and Event-triggered Tasks tab.

    2. Click the Scaling Rules tab, and then click Create Scaling Rule in the upper-right corner.

    3. In the Create Scaling Rule dialog box, configure parameters of the scaling rule and click OK.

      The following table describes the parameters that are used in this example. For parameters that are not described in the table, use the default settings.

      Parameter

      Description

      Rule Name

      Enter Delete1.

      Rule Type

      Select Simple Scaling Rule.

      Operation

      Set the value to Remove 1 Instances.

    4. Find the Delete1 scaling rule and click Execute in the Actions column.

    5. In the Execute Scaling Rule message, click OK.

    After the scaling rule is executed, Auto Scaling removes one ECS instance from the scaling group. However, the ECS instance enters the Pending Remove state because of the ESSHookForDeleteNicAndEip lifecycle hook that is in effect before the ECS instance is removed from the scaling group. During the timeout period of the lifecycle hook, Auto Scaling notifies OOS to execute the O&M operations that are defined in the ACS-ESS-LifeCycleDetachNetworkInterfaceAndDeleteEip public template.

  7. Check whether the ENI is unbound from the ECS instance and the EIP of the ENI is released.

    You can go to the Network Interfaces page to check whether the ENI exists and the Elastic IP Addresses page to check whether the EIP exists. If the ENI and the EIP do not exist, the template takes effect.

    If the ECS instance is removed from the scaling group but the ENI or the EIP still exists, you can go to the OOS console to check the execution of the O&M operations. For more information, see Step 3: (Optional) View the OOS execution.

Step 3: (Optional) View the OOS execution

  1. Log on to the OOS console.

  2. In the left-side navigation pane, choose Automated Tasks > Task Execution Management.

  3. Find the execution task by time and click Details in the Actions column.

  4. On the execution details page, view information about the OOS execution.

    For example, in the Basic Information section, you can view the execution ID and status. In the Execution Result section, you can click a task node to view the execution details. For more information, see View the details of an execution.

    Note

    If an execution failed, the error message is displayed in the Execution Result section.

FAQ

If you fail to execute an O&M task, troubleshoot the failure based on the error message in the execution result. The following section describes the common error messages and solutions:

  • Error message: Forbidden.Unauthorized message: A required authorization for the specified action is not supplied.

    Solution: Check whether the required permissions, such as the sample permissions in Step 1, are granted to the RAM role OOSServiceRole. Before OOS can manage the resources that are declared in the OOS template, you must grant required permissions to the RAM role.

  • Error message: Forbidden.RAM message: User not authorized to operate on the specified resource, or this API doesn't support RAM.

    Solution: Check whether the required permissions, such as the sample permissions in Step 1, are granted to the RAM role OOSServiceRole. Before OOS can manage the resources that are declared in the OOS template, you must grant required permissions to the RAM role.

  • Error message: LifecycleHookIdAndLifecycleActionToken.Invalid message: The specified lifecycleActionToken and lifecycleActionId you provided does not match any in process lifecycle action.

    Solution: Check whether the timeout period of the lifecycle hook is sufficient for the O&M task specified in the OOS template to complete.