You can suspend ongoing scaling processes in your scaling group, and then perform other operations. This way, you can manage scaling activities in your scaling group in a fine-grained manner. You can resume scaling processes that are suspended in your scaling group. The scaling processes that are resumed continue to run in your scaling group based on functional logic. This topic describes how to suspend and resume scaling processes. This topic also describes the effects after you suspend and resume scaling processes.

Suspend scaling processes

You can suspend ongoing scaling processes in your scaling group. After you suspend the specified scaling processes, you can perform other operations. For example, you can suspend the health check process to prevent Auto Scaling from immediately removing instances that are considered unhealthy from your scaling group. After you perform desired operations, you can manually remove the unhealthy instances from your scaling group.

When you suspend a scaling process, take note of the following items:

  • You can suspend one or more scaling processes in your scaling group. The scaling processes that are suspended in your scaling group may affect each other. For example, if you suspend the scale-in process but the health check process continues to run in your scaling group, Auto Scaling may determine that an instance is unhealthy but cannot remove the instance from your scaling group.
  • You can use the process suspension feature to manage scaling processes in your scaling group. If you want to manage instances in your scaling group, we recommend that you put the instances that you want to manage into the Standby or Protected state. For example, before you fix issues on an instance or restart an instance, you can put the instance into the Standby state. If you want to prevent an instance from being released, you can put the instance into the Protected state. For more information, see Put an ECS instance into the Standby state and Put an ECS instance into the Protected state.
  • You can suspend multiple scaling processes at the same time in your scaling group. The effect varies based on the process that you suspend. The following table describes the effects after you suspend processes.
    Scaling process Effect
    Scale-out After you suspend the scale-out process, Auto Scaling rejects all scale-out requests, which include:
    • The requests to manually add instances to the scaling group.
    • The requests to rebalance the distribution of instances in the scaling group.
    • The requests to manually execute scale-out rules or use scheduled tasks or event-triggered tasks to execute the scale-out rules when the Expected Number of Instances feature is disabled for the scaling group.
    • The requests to use event-triggered tasks to execute scale-out rules when the Expected Number of Instances feature is enabled for the scaling group.
    • The requests to add instances if you manually execute scale-out rules or use scheduled tasks to execute the scale-out rules to reach the expected number of instances in the scaling group for which the Expected Number of Instances feature is enabled. In this case, the actual number of instances remains unchanged. After the scale-out process is resumed, Auto Scaling performs checks on the expected number of instances to add instances.
    • The requests to automatically create instances when Auto Scaling performs checks on the minimum number of instances.
    • The requests to automatically create preemptible instances if the Supplemental Preemptible Instance feature is enabled.
    Scale-in After you suspend the scale-in process, Auto Scaling rejects all scale-in requests, which include:
    • The requests to manually remove instances from the scaling group.
    • The requests to rebalance the distribution of instances in the scaling group.
    • The requests to manually execute scale-in rules or use scheduled tasks to execute the scale-in rules if the Expected Number of Instances feature is disabled for the scaling group.
    • The requests to use event-triggered tasks to execute the scale-in rules.
    • The requests to remove instances if you manually execute scale-in rules or use scheduled tasks to execute the scale-in rules to reach the expected number of instances in the scaling group for which the Expected Number of Instances feature is enabled. In this case, the actual number of instances remains unchanged. After the scale-in process is resumed, Auto Scaling performs checks on the expected number of instances to remove instances from the scaling group.
    • The requests to automatically remove instances from the scaling group when Auto Scaling performs checks on the maximum number of instances.
    Health check After you suspend the health check process, Auto Scaling cannot remove instances that are considered unhealthy.
    Scheduled task After you suspend the scheduled task in the scaling group, the scaling rules that are associated with the scheduled task cannot be executed.
    Event-triggered task After you suspend the event-triggered task, the scaling rules that are associated with the event-triggered task cannot be executed if the event-triggered task enters the alert state.

Resume scaling processes

You can resume scaling processes that are suspended in your scaling group. The scaling processes that are resumed continue to run in your scaling group based on functional logic. For example, after you resume the health check process in your scaling group, Auto Scaling automatically removes instances that are considered unhealthy from the scaling group.

When you resume a scaling process, take note of the following items:

  • After a scaling process is resumed, the process may take into account the changes made during the suspension. For example, the expected number of instances in a scaling group changes when the scale-out process is suspended for the scaling group. After the scale-out process is resumed, Auto Scaling performs checks on the expected number of instances to add instances, and the expected number of instances changes based on the check result.
  • You can resume multiple scaling processes at the same time in your scaling group. The effect varies based on the process that you resume. The following table describes the effects after you resume processes.
    Scaling process Effect
    Scale-out After you resume the scale-out process, Auto Scaling continues to receive and process the scale-out requests. For example, you can manually add instances to the scaling group, and Auto Scaling can perform checks on the expected and minimum numbers of instances in the scaling group to add instances.

    If the Expected Number of Instances feature is enabled and the expected number of instances increases during the suspension, Auto Scaling performs checks on the expected number of instances to add instances.

    Scale-in After you resume the scale-in process, Auto Scaling continues to receive and process scale-in requests. For example, you can manually remove instances from the scaling group and Auto Scaling can perform checks on the expected and maximum numbers of instances to decide whether to remove instances from the scaling group.

    If the Expected Number of Instances feature is enabled and the expected number of instances decreases during a suspension, Auto Scaling performs checks on the expected number of instances to remove instances.

    Health check After you resume the health check process, Auto Scaling automatically removes instances that are considered unhealthy.
    Scheduled task If a scheduled task is not due or is being retried, the scaling rules that are associated with the task are triggered after you resume the scheduled task.
    Event-triggered task If an event-triggered task enters the alert state, the scaling rules that are associated with the task are triggered after you resume the event-triggered task.

Procedure

  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 a region.
  4. On the Scaling Groups page, find the scaling group that you want to manage and click Edit in the Actions column.
  5. In the Edit Scaling Group dialog box, perform the following operations to suspend or resume processes in the scaling group:
    • Suspend scaling processes: Select one or more processes that you want to suspend from the Suspend Process drop-down list.
    • Resume scaling processes: Remove one or more processes that you want to resume from the Suspend Process text box.
  6. Click OK.
    If the scaling strength of your scaling group is not strong, a message that prompts you to confirm your operation appears. You can confirm the operation or go back to the Edit Scaling Group dialog box to modify the parameter settings based on your business requirements.
    Note If the scaling strength of your scaling group is not strong, the scaling activities in the scaling group may fail. To prevent scaling failures, we recommend that you go back to the Edit Scaling Group dialog box to increase the scaling strength of your scaling group.
    • If you do not want to change the parameter settings, click Continue.
    • If you want to change the parameter settings, click Back to Modify. After you change the parameter settings, click OK.
  7. In the message that shows the operation is successful, click OK.

Results

On the Scaling Groups page, find the scaling group that you edited and click Details in the Actions column. On the Basic Information tab, you can view the processes that are suspended or resumed in the Scaling Group Basic Information section.

References