Configure the automatic performance scaling feature for an ApsaraDB RDS for MySQL instance - ApsaraDB RDS

Continuous business development, data volume increase, and new service or feature release can cause the database traffic to increase and computing resources to be insufficient. If the issues occur, you can increase the computing resources of your RDS instance, such as CPU cores and memory capacity. Database Autonomy Service (DAS) provides the automatic performance scaling feature for your ApsaraDB RDS for MySQL instance to handle traffic spikes during peak hours. When traffic decreases, DAS automatically scales down the instance specifications to prevent resource waste.

Prerequisites

The RDS instance uses the subscription or pay-as-you-go billing method.
Note
If the RDS instance uses the serverless billing method, the system automatically scales the storage capacity of the RDS instance. You do not need to configure automatic performance scaling.
The RDS instance runs RDS High-availability Edition.
The RDS instance uses a general-purpose instance type.
The RDS instance uses the standard product type.
The region in which the RDS instance resides supports the anomaly detection feature of DAS. For more information, see Anomaly detection.
The balance in your Alibaba Cloud account is sufficient.

Note

If the RDS instance uses a phased-out instance type and cloud disks, the automatic performance scaling feature is not supported. To use the feature, you must change the instance type to an available instance type. For information about how to change the instance type of an RDS instance, see Change instance specifications.

Billing rules

Local disks: By default, the automatic performance scaling feature is disabled, and no fees are generated. If you enable the feature, you are charged based on the pay-as-you-go billing method.

The unit price of the automatic performance scaling feature varies based on the region. The following table describes the unit prices in different regions.

Region	Unit price (USD per CPU core-hour)
China (Zhangjiakou) and China (Ulanqab)	0.063
China (Hong Kong) and South Korea (Seoul)	0.134
Japan (Tokyo)	0.100
Malaysia (Kuala Lumpur)	0.102
Singapore and Indonesia (Jakarta)	0.155
Germany (Frankfurt) and UK (London)	0.078
US (Virginia) and US (Silicon Valley)	0.129
UAE (Dubai)	0.091
Other regions	0.083

For RDS instances that use local disks, the automatic performance scaling feature is billed per minute, and a bill is generated once every hour on the hour. The fee for an automatic performance scaling operation is calculated by using the following formula: Hourly fee charged per CPU core × Number of CPU cores that are added × Scaling duration measured in hours.

For example, an RDS instance resides in the China (Hangzhou) region and provides 4 CPU cores. If the RDS instance automatically scales up to 8 CPU cores and the scaling duration spans 30 minutes, you are charged USD 0.166 based on the following calculation: USD 0.083 (the unit price per CPU core-hour) × 4 (the number of CPU cores that are added) × 0.5 (the scaling duration measured in hours) = USD 0.166. The unit price in the China (Hangzhou) region is USD 0.083 per CPU core-hour.

Cloud disks: By default, the automatic performance scaling feature is disabled, and no fees are generated. If you enable the feature, you are charged based on the new instance type that is selected during automatic performance scaling. The scaling pricing varies based on the instance configurations, such as the region and the new specifications. For more information, visit the ApsaraDB RDS buy page.

Feature description

流程图

Item	RDS instance that uses cloud disks	RDS instance that uses local disks
Expandable items	Instance type, which specifies the number of cores and the memory capacity IOPS Maximum number of connections Note When the conditions for automatic scale-ups are met, DAS automatically computes appropriate specifications for the RDS instance. This process can continue until the maximum specifications that are allowed are reached. After an automatic scale-up is complete, the IOPS and the maximum number of connections simultaneously increase. For more information about the IOPS and maximum number of connections that are supported by each instance type, see Primary ApsaraDB RDS instance types.	CPU IOPS Note The number of cores is increased by two times than the original number of cores and can be increased only once. After the number of cores is increased during an automatic scale-up, the number of cores can no longer be increased in subsequent possible automatic scale-ups. The IOPS increases by 1,000 for every core that is added.
Scale-down	Automatic scale-downs are supported. You cannot configure automatic sale-down settings for automatic performance scaling on the Basic Information page. Instead, you can choose Autonomy Services > Automatic Performance Scaling to configure automatic scale-down settings. For more information, see Use the automatic performance scaling feature.	Automatic scale-downs are supported. Note If the CPU utilization remains lower than 30% during 99% of the scale-down observation window, DAS automatically scales down the number of cores and IOPS to the original specifications.
Availability	An instance switchover occurs. Important An instance switchover may occur during the period of time for the specification change to take effect. Make sure that your application is configured to automatically reconnect to the RDS instance. For more information about the impacts of an instance switchover, see Impacts of an instance switchover. If you enable persistent connections of the database proxy feature, the connection between your application and the database proxy remains valid during an automatic switchover. For more information, see Configure persistent connection settings and Enable the database proxy feature.	No instance switchovers occur. Note An automatic scale-up or an automatic scale-down can be completed within 30 seconds. No transient connections occur during the scale-up or scale-down process.

Limits

For RDS instances that use local disks, the automatic performance scaling feature can scale up only the number of CPU cores and the IOPS and does not scale up the memory capacity or the maximum number of connections.
RDS instances that use local disks cannot automatically scale up if the hosts on which the instances are deployed cannot provide sufficient resources. The probability of a host failing to provide sufficient resources is less than 1%.
If your RDS instance uses local disks, scale-ups are performed only on the primary RDS instance. If a primary/secondary switchover is triggered after the primary RDS instance is scaled up, one of the following situations occurs:
- If the primary RDS instance (original secondary RDS instance) meets the conditions to trigger a scale-up or a scale-down, DAS performs the scaling operation on the primary RDS instance.
- If the secondary RDS instance (original primary RDS instance) meets the conditions to trigger a scale-down, the system scales the secondary RDS instance down to the original specifications.

Impacts

When the new specifications are being applied to an RDS instance, an instance switchover occurs. Make sure that your application is configured to automatically reconnect to the RDS instance. For more information about the impacts of an instance switchover, see Impacts of an instance switchover.
If an RDS instance does not run the latest minor engine version, DAS updates the minor engine version of the RDS instance to the latest version during a specification change. This ensures the performance and stability of the RDS instance.
After you enable the automatic performance scaling feature for an RDS instance, the system assigns the AliyunServiceRoleForDAS service-linked role to DAS. This way, DAS can access the resources that are related to ApsaraDB RDS.

Procedure

Log on to the ApsaraDB RDS console and go to the Instances page. In the top navigation bar, select the region in which the RDS instance resides. Then, find the RDS instance and click the ID of the instance.
In the Configuration Information section of the page that appears, click Settings next to Automatic Performance Scaling.

In the dialog box that appears, configure the following parameters.

Instance category	Parameter	Description
General-purpose RDS instance	Automatic Resource Scalability	Specifies whether to enable the automatic performance scaling feature.
General-purpose RDS instance	CPU Trigger Threshold	The average CPU utilization threshold in percentage based on which DAS scales up the RDS instance. Unit: percentage (%). Valid values: Valid values for an RDS instance that uses local disks: 70 to 90. Valid values for an RDS instance that uses cloud disks: 50 to 90 You can adjust the threshold at a step size of 10%.
RDS instance that uses cloud disks	Observation Window	The time range during which you want to observe the CPU utilization of the RDS instance. Valid values: 5 minutes, 20 minutes, 30 minutes, 40 minutes, and 60 minutes Note DAS periodically checks the CPU utilization of the RDS instance during the observation window. If the CPU utilization reaches the value of the CPU Trigger Threshold parameter, an automatic scale-up is triggered.
RDS instance that uses cloud disks	Upper Limit of Specifications	The maximum specifications that are supported by the automatic performance scaling feature. If the average CPU utilization of the RDS instance reaches the value of the CPU Trigger Threshold parameter over the specified observation window, DAS scales up your RDS instance based on your business requirements. If the average CPU utilization of the RDS instance reaches the value of the CPU Trigger Threshold parameter again over the specified observation window, DAS scales up the RDS instance again. This process continues until the specifications of the RDS instance reach the maximum specifications that are allowed. Important If the RDS instance uses an instance type that provides 1 CPU core and 2 GB of memory and an automatic scale-up is triggered, the specifications of the RDS instance are upgraded to 4 CPU cores and 8 GB of memory. In this case, you must configure the maximum specifications to which the RDS instance can be scaled up to 4 CPU cores and 8 GB of memory or above. Otherwise, the scale-up fails.
RDS instance that uses local disks	Scale-up Observation Window	Select the time range during which you want to observe the CPU utilization of the RDS instance and trigger an automatic scale-up. Valid values: 1 minute, 3 minutes, 5 minutes, 10 minutes, 20 minutes, and 30 minutes Note The system periodically checks the CPU utilization of the RDS instance during the scale-up observation window. If the CPU utilization reaches the value of the CPU Trigger Threshold parameter, an automatic scale-up is triggered.
RDS instance that uses local disks	Scale-down Observation Window	Select the time range during which you want to observe the CPU utilization of the RDS instance and trigger an automatic scale-down. Valid values: 1 minute, 3 minutes, 5 minutes, 10 minutes, 20 minutes, and 30 minutes Note DAS periodically checks the CPU utilization of the RDS instance during the scale-down observation window. If the CPU utilization remains below 30% for 99% of the scale-down observation window, an automatic scale-down is triggered.

Click OK.
In the Configuration Information section of the page, the status of Automatic Performance Scaling changes to Enabled.