How do I manually change instance states? - Auto Scaling

This topic describes how to manually put an instance into the Standby, Protected, or Stopped state, and how to move it out of the Standby and Protected states.

Instance states

Instances in a scaling group can be in one of the following service states: Adding, Pending Add, In Service, Standby, Protected, Removing, Pending Remove, and Stopped. For more information about the states of instances in a scaling group, see Lifecycle states of instances in a scaling group.

The following table describes the manual state changes for instances in a scaling group.

Instance type	Initial state	Actions	Final state	Console	API
ECS instance or elastic container instance (ECI)	Protected	Move out of Protected	In Service	Manually put an instance into or move it out of the Protected state	SetInstancesProtection
ECS instance or elastic container instance (ECI)	In Service	Put into Protected	Protected		SetInstancesProtection
ECS instance	In Service, Standby, or Protected	Put into Stopped	Stopped Note While an instance is being put into the Stopped state, it is in the Removing state.	Manually put an instance into the Stopped state	None
ECS instance or elastic container instance (ECI)	In Service, Stopped, or Protected	Delete instance or remove from scaling group	Stateless Note While an instance is being removed or deleted, it is in the Removing state.	Manually remove or delete an instance	RemoveInstances DetachInstances
ECS instance or elastic container instance (ECI)	In Service	Put into Standby	Standby	Manually put an instance into or move it out of the Standby state	EnterStandby
ECS instance or elastic container instance (ECI)	Standby	Move out of Standby	In Service		ExitStandby
ECS instance or elastic container instance (ECI)	Standby, In Service, or Protected	remove from scaling group	The removed instance is not stopped or released and exists independently of the scaling group.	Manually remove or delete an instance	DetachInstances
managed instance	In Service	remove from scaling group	The removed managed instance is not stopped or released and exists independently of the scaling group.	Manually remove a managed instance from a scaling group	None

Notes

Note the following effects when you put an ECS instance or an elastic container instance (ECI) into the Standby, Protected, or Stopped state, or move it out of the Standby state.

Stage	Description
After an instance is put into the Standby state	If the instance's scaling group is configured with a Server Load Balancer (SLB) instance, the instance's load balancing weight is set to 0. An instance remains in the Standby state until you manually move it out of this state. Auto Scaling no longer manages the instance's lifecycle. During scaling activities, Auto Scaling does not remove instances in the Standby state from the scaling group. Stopping or restarting an instance in the Standby state does not update its health check status. To release an instance in the Standby state, you must first remove it from the scaling group. If you delete the instance's scaling group, the instance is removed from the group and automatically exits the Standby state. Note If the scaling group manages the entire lifecycle of the instance, the instance is released. Otherwise, the instance is not released and continues to run. For more information about how to manage the instance lifecycle, see Instance lifecycle in a scaling group. You can stop the ECS instance or perform other operations, such as stopping an instance, restarting an instance, changing the instance type of an ECS instance, changing the operating system (replacing the system disk), re-initializing a cloud disk, and migrating an instance from the classic network to a virtual private cloud (VPC).
After an instance is moved out of the Standby state	The ECS instance or elastic container instance (ECI) returns to the In Service state. If the scaling group is associated with an SLB instance, the instance's weight is reset to the default of 50 when it is re-added to the SLB instance. Stopping or restarting the instance updates its health check status. During a scale-in event, Auto Scaling can remove the instance from the scaling group. Note If the scaling group manages the entire lifecycle of the instance, Auto Scaling releases the instance. Otherwise, the instance continues to run. For more information about how to manage the instance lifecycle, see Instance lifecycle in a scaling group.
After an instance is put into the Protected state	If the scaling group is configured with an SLB instance, the load balancing weight of the instance is not affected. The instance remains in the Protected state until you manually remove it from that state. An instance in the Protected state is not removed during scale-in events. To release it, you must first manually remove it from the scaling group. Stopping or restarting the instance does not update its health check status.
After an instance is put into the Stopped state	When an ECS instance in a scaling group enters the Stopped state, it is stopped and no longer serves traffic. You are no longer charged for some of its resources. For example, you do not need to manually enable Economical Mode for the pay-as-you-go instance in the ECS console. The associated vCPUs, memory, and fixed public IP address are released, and charges for them stop. However, other resources such as cloud disks and elastic IP addresses (EIPs) are retained and continue to incur charges.

Enter or exit Standby state

You can place temporarily idle ECS or elastic container instances (ECI) into the Standby state, and later move them back into service.

Log on to the Auto Scaling console.
In the navigation pane on the left, click Scaling Groups.
In the top navigation bar, select a region.
Find the target scaling group and click View Details in the Actions column.
At the top of the page, click the Instance Management tab.
Find the instance in the appropriate list.
- To select an automatically created instance, go to the Auto Created list.
- To select a manually added instance, go to the Manually Added list.

Manually put an instance into or move it out of the Standby state.

Enter Standby

Use one of the following methods to put an instance into the Standby state:
- Single instance: Find the target instance and click Switch to Standby State in the Actions column.
- Multiple instances: Select the target instances and click Switch to Standby State at the bottom of the instance list.

In the Switch to Standby State dialog box, choose an operation method.

Actions

Description

Procedure

Synchronous operation

The task is processed in the background after you submit it. This can cause long waits or timeouts for scaling groups with many resources. We recommend choosing the asynchronous method in such scenarios.

Clear the Asynchronously Switch to Standby checkbox and click OK.

Asynchronous operation

The console immediately confirms task submission, providing a better user experience than a synchronous operation. You can view the final execution status on the Scaling Activities tab.

Select the Asynchronously Switch to Standby checkbox and click OK.

On the Scaling Rules and Activities tab, click the Scaling Activities tab to view the execution status.

Exit Standby

Use one of the following methods to move an instance out of the Standby state:
- Single instance: Find the target instance and click Remove from Standby in the Actions column.
- Multiple instances: Select the target instances and click Remove from Standby at the bottom of the instance list.

In the Remove from Standby dialog box, choose an operation method.

Actions

Description

Procedure

Synchronous operation

Clear the Asynchronously Remove from Standby checkbox and click OK.

Asynchronous operation

The console immediately confirms task submission, providing a better user experience than a synchronous operation. You can view the final execution status on the Scaling Activities tab.

Select the Asynchronously Remove from Standby checkbox and click OK.

On the Scaling Rules and Activities tab, click the Scaling Activities tab to view the execution status.

Enter or exit Protected state

You can put an ECS or elastic container instance (ECI) into the Protected state to prevent its removal from a scaling group, and you can also move it out of this state.

Log on to the Auto Scaling console.
In the navigation pane on the left, click Scaling Groups.
In the top navigation bar, select a region.
Find the target scaling group and click View Details in the Actions column.
At the top of the page, click the Instance Management tab.
Find the instance in the appropriate list.
- To select an automatically created instance, go to the Auto Created list.
- To select a manually added instance, go to the Manually Added list.
Manually put an instance into or move it out of the Protected state.
Enter Protected
1. Use one of the following methods to put an instance into the Protected state:
  - Single instance: Find the target instance and click Switch to Protected State in the Actions column.
  - Multiple instances: Select the target instances and click Switch to Protected State at the bottom of the instance list.
2. Click OK.
Exit Protected
1. Use one of the following methods to move an instance out of the Protected state:
  - Single instance: Find the target instance and click Remove from Protected State in the Actions column.
  - Multiple instances: Select the target instances and click Remove from Protected State at the bottom of the instance list.
2. Click OK.

Put an instance into Stopped state

This operation applies only to ECS scaling groups in a Virtual Private Cloud (VPC). If the release mode for instances in the scaling group is set to Economical Mode and these instances are automatically created ECS instances, you can manually stop them. When a scale-out event occurs, Auto Scaling preferentially starts the Stopped ECS instances.

Log on to the Auto Scaling console.
In the navigation pane on the left, click Scaling Groups.
In the top navigation bar, select a region.
Find the target scaling group and click View Details in the Actions column.
At the top of the page, click the Instance Management tab.
Find the Auto Created list.
Find the target ECS instance. In the Actions column, choose >Stop Instance.
Click OK.

Remove managed instances

This operation applies only to scaling groups that contain ECS instances. If you no longer need a scaling group to manage a managed instance, you must manually remove it. The removed managed instance is not released.

Log on to the Auto Scaling console.
In the navigation pane on the left, click Scaling Groups.
In the top navigation bar, select a region.
Find the target scaling group and click View Details in the Actions column.
At the top of the page, click the Instance Management tab.
Find the Managed Instance list.
Use one of the following methods to remove managed instances from the scaling group:
- Single instance: Find the target managed instance and click Remove from Scaling Group in the Actions column.
- Multiple instances: Select the target managed instances and click Remove from Scaling Group at the bottom of the list.
Choose whether to Modify Expected Number of Instances (selected by default), and then click OK.
- If you select Modify Expected Number of Instances, the expected number of instances decreases by the number of removed instances.
- If you clear Modify Expected Number of Instances, the expected number of instances does not change after you remove the instances.