Starting July 5, 2024, ApsaraDB RDS for MySQL instances that run RDS High-availability Edition with local disks and use general-purpose instance types support automatic performance scaling, powered by Database Autonomy Service (DAS).
How it works
Traffic spikes and sudden workload surges put enormous pressure on database CPU. Without automation, responding to a spike requires manually triggering a scale-up through the console, waiting an unpredictable amount of time for the operation to complete, and then manually scaling back down after traffic normalizes. This process is slow, disruptive, and error-prone.
Automatic performance scaling eliminates this manual cycle. When DAS detects that CPU utilization is approaching its limit, it automatically adds CPU cores and increases IOPS. When traffic drops, it scales back down to the original specifications to avoid unnecessary costs.
Effective date
July 5, 2024
Billing
Automatic performance scaling is disabled by default and incurs no charges. When enabled, it uses pay-as-you-go billing, charged per minute and billed once every hour on the hour.
Formula: Unit price (USD/CPU core-hour) × Number of CPU cores added × Scaling duration (hours)
Example: An instance in the China (Hangzhou) region scales from 4 cores to 8 cores for 30 minutes. China (Hangzhou) falls under the "Other regions" pricing tier.
Unit price: USD 0.083/CPU core-hour
Cores added: 4
Duration: 0.5 hours
Total: USD 0.083 × 4 × 0.5 = USD 0.166
Regional unit prices:
| 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 |
| Australia (Sydney) Closing Down | 0.106 |
| Other regions | 0.083 |
Prerequisites
Automatic performance scaling is available for instances that meet all of the following criteria:
Billing method: subscription or pay-as-you-go
Serverless instances scale automatically. Automatic performance scaling does not apply to serverless instances.
RDS edition: RDS High-availability Edition
Storage type: local disk
Product type: standard
Instance type: general-purpose
Region: the region supports the anomaly detection feature of DAS
Account: sufficient balance in your Alibaba Cloud account
Limitations
What scales and what doesn't
When automatic performance scaling triggers, it increases CPU cores and IOPS. Memory capacity and the maximum number of connections remain unchanged.
Host resource availability
If the host cannot provide enough resources for a scale-up, the operation does not proceed. The probability of this occurring is less than 1%.
Primary instance only
Scaling applies only to primary RDS instances. Secondary instances are not scaled independently.
Behavior after a primary/secondary switchover
If a primary/secondary switchover occurs after a scale-up:
If the new primary instance (original secondary) meets the conditions for a scale-up or scale-down, the system performs the corresponding scaling operation.
If the new secondary instance (original primary) meets the conditions for a scale-down, the system scales it down to the original specifications.