This topic describes the per-second billing model for pay-as-you-go AnalyticDB for PostgreSQL instances that is in effect starting from August 2, 2022.

Description

The per-second billing model is in effect for pay-as-you-go AnalyticDB for PostgreSQL instances. AnalyticDB for PostgreSQL instance resources are billed in one-second increments, and a bill is generated for fee deduction at the end of each settlement period. If the specifications of an instance are changed or an instance is paused and resumed within a settlement period, the instance is billed in one-second increments based on the current specifications or state.

Applicable resources

  • Elastic storage mode
    • Computing resources
    • Storage resources
  • Serverless mode
    • Computing resources
    Note In Serverless mode, storage resources are still billed on an hourly basis.

Settlement period

Each settlement period starts and ends on the hour. A bill is generated at the end of each settlement period. For more information about bills, see Bills.

For example, assume that you created a pay-as-you-go instance at 10:59:30 and then released it at 12:50:30. Bills are generated during the following settlement periods:

  • The instance is created at 10:59:30, which is within the settlement period that is from 10:00:00 to 11:00:00. Its billing starts from the second it is created till 11:00:00. The billing duration is 30 seconds.
  • The instance keeps running in the settlement period that is from 11:00:00 to 12:00:00. It is billed for all the 3,600 seconds of the settlement period.
  • In the next settlement period from 12:00:00 to 13:00:00, the billing of the instance continues until 12:50:30 when the instance is released. The billing duration is 3,030 seconds.

Billing changes when instance specifications are changed

When the specifications of an instance are changed, the instance goes through three states: Running, Scaling, and Running. When the instance is in the first Running state or the Scaling state, the instance resources are billed based on the original specifications. When the instance is in the second Running state, the instance resources are billed based on the new specifications.

For example, assume that you changed the specifications of a pay-as-you-go instance and the change was complete at 11:30:00. The following bills are generated during the 11:00:00 to 12:00:00 settlement period:

  • The first bill is generated for the period of 11:00:00 to 11:30:00 based on the instance specifications before the change.
  • The second bill is generated for the period of 11:30:00 to 12:00:00 based on the instance specifications after the change.

Billing changes when instances are paused and resumed

When an instance is paused and resumed, the instance goes through five states: Running, Pausing, Paused, Starting, and Running. When the instance is in the Paused or Starting state, no charges are incurred.

For example, assume that you paused an instance and the instance state changed to Paused at 11:20:00. After a while, you resumed the instance and the instance state changed to Running at 11:40:00. The following bills are generated during the 11:00:00 to 12:00:00 settlement period:

  • The first bill is generated for the period of 11:00:00 to 11:20:00 based on the current instance specifications.
  • No bills are generated for the period of 11:20:00 to 11:40:00, during which the instance is paused and not running.
  • The second bill is generated for the period of 11:40:00 to 12:00:00 based on the current instance specifications.
Note For instances in Serverless mode, the preceding rules apply only to computing resources, while storage resources are still billed on an hourly basis.