This topic describes the release notes for Auto Scaling and provides links to the relevant references.
Release notes in 2024
February 2024
Feature | Description | Release date | Region | References |
Support for LOAD_BALANCER | The | 2024-02-28 | All regions | |
Support for management of scaling groups and scaling configurations based on Kubernetes Deployment files in YAML format | Kubernetes Deployment files in the YAML format are supported to create and manage scaling groups and scaling configurations of the Elastic Container Instance type. | 2024-02-27 | All regions | - |
Support for the Entrusted parameter | The | 2024-02-22 | All regions | - |
Support for restarting instances from the Economical Mode | The | 2024-02-22 | All regions | |
Support for changing the instance reclaim mode | The | 2024-02-22 | All regions |
January 2024
Feature | Description | Release date | Region | References |
Support for custom priority policies in scaling configurations of the ECS type | The | 2024-01-29 | All regions | |
Support for lifecycle management of containers | Lifecycle management of containers is supported. After you enable this feature, you can use methods such as command scripts, HTTP requests, and TCP connections to probe containers. | 2024-01-19 | All regions |
Release notes in 2023
December 2023
Feature | Description | Release date | Region | References |
Support for the Forcibly Recycle setting | The Forcibly Recycle setting is supported by scaling groups of the ECS type. If you set the Instance Reclaim Mode parameter for a scaling group to | 2023-12-14 | All regions | |
Support for balanced distribution of elastic container instances across regions | Balanced distribution of instances is supported by scaling groups of the Elastic Container Instance type. This feature is automatically enabled and does not require manual configuration. When a scale-out is triggered in a scaling group, Auto Scaling balances the distribution of newly created elastic container instances across regions based on the current instance distribution. | 2023-12-13 | All regions |
November 2023
Feature | Description | Release date | Region | References |
Support for the DataCache feature | The DataCache feature is supported by scaling configurations of the Elastic Container Instance type. You can enable this feature to pull data in advance and save cached data to your on-premises computer. This way, you can mount the cached data to elastic container instances when you create the elastic container instances. This helps reduce instance startup time, prevent repeated data downloads, and save resource costs. | 2023-11-15 | All regions |
October 2023
Feature | Description | Release date | Region | References |
Support for the ClientToken parameter | The | 2023-10-18 | All regions | |
Support for disabling the Expected Number of Instances feature | The | 2023-10-8 | All regions |
September 2023
Feature | Description | Release date | Region | References |
Support for the Forcibly Release setting | The Forcibly Release setting is supported. When you create a scaling group, you can set the Instance Reclaim Mode parameter to Forcibly Release. In this case, Auto Scaling forcibly releases the instances that are in the Running state during scale-ins. | 2023-09-28 | All regions | |
Support for ECS instance types in scaling configurations of the Elastic Container Instance type | ECS instance types are supported in scaling configurations of the Elastic Container Instance type. You can use the InstanceTypes parameter to specify multiple ECS instance types to create elastic container instances. | 2023-09-28 | All regions |
August 2023
Feature | Description | Release date | Region | References |
Support for instance release protection | The DeletionProtection parameter is supported by scaling configurations of the ECS type. If you set the DeletionProtection parameter to true when you call an API operation or use the Auto Scaling console to create a scaling configuration of the ECS type, the instance release protection feature is enabled. In this case, you cannot release pay-as-you-go ECS instances by calling an API operation or using the Auto Scaling console. | 2023-08-16 | All regions |
July 2023
Feature | Description | Release date | Region | References |
Support for management of Alibaba Cloud instances and non-Alibaba Cloud instances | You can use Cloud Assistant to register a non-Alibaba Cloud instance as an Alibaba Cloud-managed instance and install a CloudMonitor agent on the registered instance. Then, you can add the registered instance to a scaling group. This way, you can monitor non-Alibaba Cloud instances and Alibaba Cloud instances in a centralized manner, and instances are automatically scaled based on metric statistics. | 2023-07-03 | All regions |
June 2023
Feature | Description | Release date | Region | References |
Support for event bus as the notification method | The event bus feature of EventBridge is supported. If you use an event bus as the notification method for lifecycle hooks, Auto Scaling sends notifications of the | 2023-06-26 | All regions |
May 2023
Feature | Description | Release date | Region | References |
Support for the ECS Current Connections metric | The ECS Current Connections metric is supported. When you create an event-triggered task of the system monitoring type for a scaling group of the ECS type, you can specify the (ECS) ECS Current Connections metric. This metric is used to monitor the current number of network connections to ECS instances in the scaling group, which helps you better understand the network load of the ECS instances. | 2023-05-25 | All regions |
April 2023
Feature | Description | Release date | Region | References |
Support for billing diagnosis and inventory diagnosis | Billing diagnosis and inventory diagnosis are supported. The billing diagnosis feature allows you to check whether your Alibaba Cloud account has overdue payments. The inventory diagnosis feature allows you to check whether the instance types that are specified in your scaling configuration have sufficient resources. Billing diagnosis and inventory diagnosis help you detect and resolve issues at the earliest opportunity. | 2023-04-18 | All regions | |
Support for sequential and unique instance names or hostnames | Naming rules to generate sequential and unique instance names or hostnames are supported in scaling configurations. Auto Scaling generates sequential and unique instance names or hostnames for ECS instances that are created during scale-outs based on the naming rules. You can use the instance names or hostnames to manage ECS instances in an efficient manner. | 2023-04-15 | All regions | |
Support for logon to ECS instances by using ecs-user | ecs-user can be used to log on to ECS instances that are created by Auto Scaling during scale-outs. To use ecs-user, set the ImageOptions.LoginAsNonRoot parameter to Yes when you call an API operation to create a scaling configuration of the ECS type. | 2023-04-12 | All regions | |
Support for the weight setting of backend server groups | The weights of backend server groups of the associated Server Load Balancer (SLB) instances can be specified. When you create a scaling group, you can specify weights for instances that are added as backend servers to the backend server groups. When you create a scaling configuration, you can also configure the SLB Weight parameter to specify weights for instances that are added as backend servers. | 2023-04-10 | All regions | |
Support for the (ALB) QPS per Backend Server metric | The (ALB) QPS per Backend Server metric can be used to create even-triggered tasks and enable automatic scaling activities. This improves high availability of your applications. | 2023-04-03 | All regions | Use (ALB) QPS per Backend Server to trigger scaling activities |
March 2023
Feature | Description | Release date | Region | References |
Support for the health diagnosis feature | The health diagnosis feature of Auto Scaling is supported. You can use the health diagnosis feature to check the health status of your scaling group in a comprehensive manner. This way, you can detect exceptions in your scaling group at the earliest opportunity and resolve common issues. This improves the success rate of your scaling activities. | 2023-03-21 | All regions |
February 2023
Feature | Description | Release date | Region | References |
Association with Network Load Balancer (NLB) instances | NLB instances can be associated with scaling groups. After you associate an NLB instance with a scaling group, Auto Scaling adds the ECS instances or elastic container instances that are added to the scaling group as backend servers of the associated NLB instance. This way, the NLB instance can forward access requests to backend servers in an efficient manner. | 2023-02-03 | All regions |
January 2023
Feature | Description | Release date | Region | References |
Update of the Containers settings | The Containers settings in scaling configurations that are used to create elastic container instances can be updated in full update and incremental update modes. You can select an update mode based on your business requirements. | 2023-01-03 | All regions | |
Query of the image source | The image source of a scaling configuration can be queried. You can call the | 2023-01-10 | All regions |
Release notes in 2022
December 2022
Feature | Description | Release date | Region | References |
Addition of scaling groups to resource groups | Scaling groups can be added to resource groups. When you create a scaling group, you can specify a resource group for the scaling group and grant specific permissions to the resource group. This way, you can manage your scaling groups in a fine-grained manner. | 2022-12-27 | All regions |
November 2022
Feature | Description | Release date | Region | References |
Support for the | The | 2022-11-02 | All regions | |
Support for the SyncActivity parameter | The | 2022-11-02 | All regions |
October 2022
Feature | Description | Release date | Region | References |
Support for custom scale-in policies (Function) | The CustomPolicy setting is supported for the RemovalPolicy parameter. The CustomPolicyARN parameter is supported to specify custom scale-in policies. You can create custom scale-in policies by using functions to remove ECS instances from a scaling group. This allows you to remove instances from scaling groups in a more flexible manner. | 2022-10-14 | All regions | |
Support for disabling or enabling lifecycle hooks | Lifecycle hooks can be enabled or disabled. You can enable or disable lifecycle hooks in the Auto Scaling console or by calling the ModifyLifecycleHook operation. Lifecycle hooks that are in the Disabled state do not take effect. We recommend that you enable or disable lifecycle hooks based on your business requirements. | 2022-10-13 | All regions |
August 2022
Feature | Description | Release date | Region | References |
Support for the maximum life span feature of ECS instances | The Maximum Life Span of Instance parameter is supported for scaling groups that contain ECS instances. If the maximum life span of an ECS instance expires, Auto Scaling creates an ECS instance to replace the expired ECS instance. | 2022-08-25 | All regions |
July 2022
Feature | Description | Release date | Region | References |
Support for the cost optimization feature of elastic container instances | The cost optimization feature is supported for scaling groups that contain elastic container instances. After you enable the cost optimization feature, Auto Scaling compares the prices of ECS instance types, and then selects the ECS instance type that is provided at the lowest price to create elastic container instances based on your business requirements. This way, the elastic container instances that host your applications can be scaled automatically and at lower costs. | 2022-07-29 | All regions | |
Support for multi-metric alert rules in event-triggered tasks | Multi-metric alert rules are supported in event-triggered tasks. In a multi-metric alert rule, you can specify whether all metrics must meet the alert conditions before an automatic scaling activity is triggered or only one of the metrics meets the alert condition before an automatic scaling activity is triggered. For example, you can specify that both the CPU utilization and the memory usage must reach the thresholds before a scaling activity is triggered. | 2022-07-27 | All regions | |
Support for ESSD AutoPL disks | The ESSD AutoPL disks are supported in scaling configurations and launch templates. You can create ECS instances that use ESSD AutoPL disks in scaling groups. ESSD AutoPL disks inherit the features and performance of enhanced SSDs (ESSDs) and provide the capability of decoupling disk capacity from disk performance. | 2022-07-14 | All regions |
June 2022
Feature | Description | Release date | Region | References |
Support for putting elastic container instances into or moving elastic container instances out of the Standby state | The EnterStandby and ExitStandby operations are supported. You can call the EnterStandby API operation to put an elastic container instance into the Standby state. You can call the ExitStandby API operation to move an elastic container instance out of the Standby state. | 2022-06-29 | All regions | |
Support for attaching elastic container instances to or detaching elastic container instances from Application Load Balancer (ALB) instances and ApsaraDB RDS instances, and support for adding elastic container instances to or removing elastic container instances from vServer groups and ALB server groups | API operations can be called to attach elastic container instances to or detach elastic container instances from ALB instances and ApsaraDB RDS instances. API operations can be called to add elastic container instances to or remove elastic container instances from one or more vServer groups or ALB server groups. | 2022-06-14 | All regions | |
Support for filter conditions in intelligent configuration mode | Filter conditions are supported in intelligent configuration mode to help you select instance types in an efficient manner. For example, you can specify the following filter conditions in intelligent configuration mode: the instance types that you want to exclude, the instance architecture, and whether to include burstable instance types. | 2022-06-08 | All regions | |
Support for the 15-second statistical period by CloudMonitor metrics | The 15-second statistical period can be applied to the | 2022-06-08 | All regions | |
Support for the ScaleWithAdjustment operation by scaling groups with an expected number of instances | The ScaleWithAdjustment operation is supported. If you specify an expected number of instances but you do not create scaling rules when you create your scaling group, you can call the ScaleWithAdjustment operation to automate scaling. | 2022-06-06 | All regions |
May 2022
Feature | Description | Release date | Region | References |
Support for lifecycle hooks on elastic container instances | Lifecycle hooks are supported on elastic container instances. You can use lifecycle hooks to manage the lifecycles of elastic container instances in a scaling group. You can use lifecycle hooks to suspend scaling activities that are triggered by Auto Scaling and put elastic container instances into the Pending Add or Pending Remove state. You can perform operations on the elastic container instances until the timeout period of your lifecycle hooks ends. | 2022-05-17 | All regions |
April 2022
Feature | Description | Release date | Region | References |
Support for the AttachInstances and DetachInstances operations by elastic container instances | The AttachInstances operation can be called to add elastic container instances to a scaling group. The DetachInstances operation can be called to remove elastic container instances from a scaling group. | 2022-04-24 | All regions | |
Support for resource-level authentication and tag-based authentication | Resource-level authentication and tag-based authentication are supported. Different Resource Access Management (RAM) users have different permissions on Auto Scaling resources. | 2022-04-13 | All regions |
February 2022
Feature | Description | Release date | Region | References |
Support for health checks on elastic container instances | Health checks can be performed on elastic container instances in a scaling group at specified intervals. This way, elastic container instances that are unhealthy can be replaced in a timely manner. If an elastic container instance in a scaling group is not in the Running state, Auto Scaling considers the instance unhealthy and creates an elastic container instance to replace the unhealthy instance. | 2022-02-24 | All regions |
Release notes in 2021
November 2021
Feature | Description | Release date | Region | References |
Support for asynchronous interaction | An ECS instance can be put into the Standby state or moved out of the Standby state in an asynchronous manner. | 2021-11-29 | All regions | |
Support for evaluation of ECS instance availability | The scaling strength feature is provided. You can use this feature to evaluate the availability of ECS instances and improve the success rate of scale-out activities. | 2021-11-28 | All regions | |
Support for automatic notification when errors occur in scaling groups | If your scaling group is abnormal, an error message is displayed in the Auto Scaling console. The error message helps you identify and fix the errors at the earliest opportunity. | 2021-11-16 | All regions | |
Support for the configuration of capacity reservation | The Capacity Reservation parameter can be configured for ECS instances or elastic container instances when you create or modify a scaling configuration. | 2021-11-02 | All regions |
October 2021
Feature | Description | Release date | Region | References |
Support for automatic scaling of elastic container instances | Elastic container instances can be automatically scaled in a scaling group. | 2021-10-20 | All regions |
September 2021
Feature | Description | Release date | Region | References |
Intelligent configuration | Support for intelligent configuration in scaling configurations Intelligent configuration is supported to automatically filter instance types that meet your scale-out requirements. | 2021-09-09 | All regions |
August 2021
Feature | Description | Release date | Region | References |
Economical mode | The feature name is changed from No Fees for Stopped Instances (VPC-Connected) to Economical Mode. | 2021-08-09 | All regions |
June 2021
Feature | Description | Release date | Region | References |
Association with ALB server groups | ALB server groups can be associated with scaling groups. After you associate an ALB server group with a scaling group, ECS instances that are added to the scaling group are automatically added to the ALB server group as backend servers. ALB instances distribute requests to the ALB server group. | 2021-06-11 | All regions |
March 2021
Feature | Description | Release date | Region | References |
Vertical scaling | The specifications of pay-as-you-go instances can be changed. The specifications include the number of vCPUs and the memory size. | 2021-03-31 | All regions |
February 2021
Feature | Description | Release date | Region | References |
Scaling configuration | Weights can be specified for instance types. | 2021-02-12 | All regions |
January 2021
Feature | Description | Release date | Region | References |
Scaling configuration | Unique sequential hostnames can be automatically generated. | 2021-01-15 | All regions |
Release notes in 2020
November 2020
Feature | Description | Release date | Region | References |
Scaling configuration | The performance levels (PLs) of system disks or data disks of the ESSD category can be specified in scaling configurations. | 2020-11-18 | All regions |
October 2020
Feature | Description | Release date | Region | References |
Event-triggered task | More statistical methods are supported in event-triggered tasks for custom metrics. | 2020-10-25 | All regions |
September 2020
Feature | Description | Release date | Region | References |
Modification of scaling configurations | The images that are specified in scaling configurations can be modified. | 2020-09-24 | All regions |
August 2020
Feature | Description | Release date | Region | References |
Rolling update | OOS packages can be used to install software. | 2020-08-20 | Some regions | |
Expected number of instances | The expected number of instances in a scaling group can be changed when instances are deleted or removed from the scaling group. | 2020-08-12 | All regions | |
Scheduled task | The expiration time can be skipped. The maximum recurrence period can be up to one year. | 2020-08-07 | All regions |
July 2020
Feature | Description | Release date | Region | References |
Modification of scaling groups | The health check feature can be enabled or disabled for instances in an existing scaling group. | 2020-07-30 | All regions | |
Image update in scaling configurations | Tasks can be created to update multiple images in scaling configurations at the same time. | 2020-07-02 | All regions |
May 2020
Feature | Description | Release date | Region | References |
Auto Scaling-linked role | The AliyunServiceRoleForAutoScaling role is automatically created when Auto Scaling is activated. | 2020-05-06 | All regions |
April 2020
Feature | Description | Release date | Region | References |
Rolling update | Rolling update tasks can be created to update the configurations of ECS instances at the same time. | 2020-04-26 | Some regions |
March 2020
Feature | Description | Release date | Region | References |
Event-triggered task of the system monitoring type | More metrics are supported by event-triggered tasks of the system monitoring type. | 2020-03-17 | All regions | |
Scaling group creation | The classification of the configuration items that are used to create a scaling group is optimized. You can add tags to scaling groups and select existing instances as templates to create instances in the scaling groups. | 2020-03-11 | All regions |
Release history
Release notes in 2019
December 2019
Feature | Description | Release date | Region | References |
Support for suspending and resuming processes in scaling groups | The processes in a scaling group can be suspended or resumed to manage scaling operations in a fine-grained manner. | 2019-12-03 | All regions |
October 2019
Feature | Description | Release date | Region | References |
Configuration of the health status of ECS instances | An API operation can be called to specify the health status of an ECS instance. | 2019-10-25 | All regions | |
Expected number of instances | The expected number of ECS instances in a scaling group can be specified. This way, Auto Scaling can execute parallel scaling activities. | 2019-10-22 | All regions |
September 2019
Feature | Description | Release date | Region | References |
Creation of a scaling configuration from an existing instance | API operations can be called to create a scaling group, and instance IDs can be specified. Auto Scaling can obtain information about the configuration of an instance and automatically creates a scaling configuration. | 2019-09-24 | All regions |
August 2019
Feature | Description | Release date | Region | References |
Association of a lifecycle hook with an OOS template | A lifecycle hook can be associated with an OOS template. When the lifecycle hook is triggered, Auto Scaling automatically executes the associated OOS template. | 2019-08-14 | All regions |
July 2019
Feature | Description | Release date | Region | References |
Support for mixed instance allocation in the cost optimization policy | If you set the multi-zone scaling policy to cost optimization, pay-as-you-go instances and preemptible instances can be created based on the specified capacity. | 2019-07-02 | All regions |
May 2019
Feature | Description | Release date | Region | References |
Step scaling rule | A step scaling rule can be created to manage scaling activities based on the size of the alert breach. | 2019-05-20 | All regions | |
Support for multiple security groups | An instance that is automatically created can be added to multiple security groups. | 2019-05-15 | All regions | |
Query of the region information | An API operation can be called to query the regions where Auto Scaling is available. | 2019-05-05 | All regions |
April 2019
Feature | Description | Release date | Region | References |
Predictive scaling rule | Predictive scaling rules can be created. After you create a predictive scaling rule, Auto Scaling adjusts the minimum and maximum numbers of instances in a scaling group based on historical monitoring data. | 2019-04-18 | All regions | |
Support for the DiskName, DiskDescription, and Encrypted parameters in scaling configurations | The DiskName, DiskDescription, and Encrypted parameters in scaling configurations can be configured. | 2019-04-04 | All regions |
January 2019
Feature | Description | Release date | Region | References |
Target tracking scaling rule | Target tracking scaling rules can be created. After you create a target tracking scaling rule, Auto Scaling adjusts the number of instances based on the specified values. | 2019-01-11 | All regions |
Release notes in 2018
December 2018
Feature | Description | Release date | Region | References |
Change of vSwitches | The vSwitches that are associated with a scaling group can be changed. | 2018-12-28 | All regions | |
Cost optimization | The multi-zone scaling policy can be set to cost optimization. | 2018-12-16 | All regions | |
vServer group | vServer groups can be specified for the SLB instance that is associated with a scaling group. | 2018-12-07 | All regions |
November 2018
Feature | Description | Release date | Region | References |
Support for the No Fees for Stopped Instances (VPC-Connected) state | ECS instances that are removed from a scaling group can enter the No Fees for Stopped Instances (VPC-Connected) state. | 2018-11-28 | All regions |
October 2018
Feature | Description | Release date | Region | References |
Enhanced IP address whitelists of ApsaraDB RDS instances | The private IP addresses of ECS instances in a scaling group can be added to the enhanced IP address whitelists of the ApsaraDB RDS instances that are associated with the scaling group. | 2018-10-19 | All regions |
September 2018
Feature | Description | Release date | Region | References |
New region | Auto Scaling is available in the UK (London) region. | 2018-09-30 | All regions | |
Shutdown and reclaim mode | The shutdown and reclaim mode can be specified for a scaling group. | 2018-09-13 | All regions | |
Launch template | A launch template can be specified as an instance configuration source for a scaling group. | 2018-09-01 | All regions |
August 2018
Feature | Description | Release date | Region | References |
Cron expression | Cron expressions can be used to specify the recurrence period of scheduled tasks. | 2018-08-20 | All regions | |
Expiration notification for scheduled tasks | Expiration notifications for scheduled tasks can be configured. | 2018-08-20 | All regions |
July 2018
Feature | Description | Release date | Region | References |
Modification of information about data disks and the price limits of preemptible instances | Information about data disks and the price limits of preemptible instances in scaling configurations can be modified. | 2018-07-27 | All regions | |
Association with ApsaraDB RDS instances | ApsaraDB RDS instances can be associated with or disassociated from scaling groups. | 2018-07-11 | All regions | |
Association with SLB instances | SLB instances can be associated with or disassociated from scaling groups. | 2018-07-04 | All regions |
June 2018
Feature | Description | Release date | Region | References |
Modification of scaling configurations | Existing scaling configurations can be modified. | 2018-06-27 | All regions | |
Lifecycle hook | If an ECS instance is automatically created for or removed from a scaling group, lifecycle hooks can be used to put the instance into the Pending Add or Pending Remove state. This way, you can perform desired operations on the instance. | 2018-06-01 | All regions |
April 2018
Feature | Description | Release date | Region | References |
Scaling activity notification | When a specific type of scaling activity occurs in a scaling group, notifications can be sent to CloudMonitor or Message Service (MNS). | 2018-04-20 | All regions |
March 2018
Feature | Description | Release date | Region | References |
Protected state | ECS instances in a scaling group can be put into the Protected state. Instances in the Protected state cannot be released or removed from the scaling group during scale-ins. | 2018-03-28 | All regions | |
Standby state | ECS instances in a scaling group can be put into the Standby state. This way, you can manage the lifecycles of the ECS instances. | 2018-03-15 | All regions |
Release notes in 2015
Feature | Description | Release date | Region | References |
Auto Scaling | Auto Scaling is officially launched. | 2015-08-27 | Some regions |
Release notes in 2014
Feature | Description | Release date | Region | References |
Auto Scaling | Auto Scaling is in internal preview. | 2014-10-15 | Some regions |