Serverless ApsaraDB RDS for PostgreSQL instances are now generally available for commercial use. Serverless instances automatically scale computing resources up and down based on your workload — so you get full performance during peak hours and pay only for what you use during off-peak hours, with no manual intervention required.
Effective date: April 10, 2023
Usage notes
-
Serverless RDS instances are available in the following regions: China (Hangzhou), China (Shanghai), China (Beijing), China (Zhangjiakou), China (Shenzhen), China (Chengdu), China (Hong Kong), and Singapore. Additional regions will be supported in the future. Check the ApsaraDB RDS buy page for the current list.
-
To create a serverless RDS instance using a RAM user, attach the AliyunRDSFullAccess policy to that RAM user. For more information, see Use RAM for resource authorization.
How it works
The following diagram compares resource utilization between a regular RDS instance and a serverless RDS instance under fluctuating workloads.
With a regular RDS instance, resources are fixed: you overpay during off-peak hours when utilization is low, and risk performance degradation during peak hours when demand exceeds capacity.
A serverless RDS instance eliminates both problems:
-
Scales with your workload — computing resources adjust automatically in response to traffic changes, keeping resource utilization high and idle capacity low.
-
Maintains performance at peak — resources scale up to meet demand during busy periods, preserving stability and throughput.
-
Charges based on actual usage — you pay for the computing resources your workload actually consumes, not a fixed allocation.
-
No operations and maintenance (O&M) required — scaling happens automatically, with no manual capacity adjustments needed.
-
Auto start and stop — when no connections are active, the instance suspends automatically to stop resource consumption. When a connection is established, the instance resumes automatically.
-
Optimized for high-throughput and high-concurrency workloads — the instance is suited for large data volumes with significant traffic spikes, supporting high-throughput write operations and high-concurrency operations.
Billing
Serverless RDS instances use pay-as-you-go billing. Charges are based on the actual computing resources consumed by your workload. For pricing details, see Pricing of serverless ApsaraDB RDS for PostgreSQL instances.