Optimization of the Memory Usage metric for an ApsaraDB RDS for PostgreSQL instance with cloud disks - ApsaraDB RDS

ApsaraDB RDS for PostgreSQL instances that use cloud disks are deployed based on cloud services. Memory must be reserved for the system and control services. The current Memory Usage metric cannot accurately measure workloads. After the optimization, the Memory Usage metric can measure workloads more accurately and help you monitor the status of your RDS instance more efficiently.

Content

The computing logic of the memory usage is optimized.

Before the optimization, the memory usage is calculated by using the following formula: Memory usage = Memory occupied by PostgreSQL processes/Memory purchased for your RDS instance
After the optimization, the memory usage is calculated by using the following formula: Memory usage = (Memory occupied by PostgreSQL processes + Memory consumed by the management system + Memory consumed by the operating system)/Memory purchased for your RDS instance

Note

After the optimization, the following items are added to calculate the actual memory usage in a more accurately manner:

The memory consumed by the management system and the memory consumed by the operating system
The memory occupied by PostgreSQL processes includes the memory that is reserved for huge pages

Effective date

November 30, 2022

Optimization results and impacts

You can view the Memory Usage (%) metric on the Autonomy Services > Dashboard page of the ApsaraDB RDS console.
The instance specifications and performance of your RDS instance are not affected. After the computing logic of the memory usage is optimized, the accuracy of the Memory Usage (%) metric is increased.
Note
For more information about how to view metrics, see Use the performance trends feature for an ApsaraDB RDS for PostgreSQL instance. You can also view the enhanced monitoring metrics to query the db.mem_available.ratio metric. This metric indicates the percentage of the available database memory.
After the computing logic for the memory usage is optimized, the available memory of some dedicated RDS instances that have small specifications may be low. In this case, we recommend that you change the instance types to general-purpose instance types. For more information, see Change the specifications of an ApsaraDB RDS for SQL Server instance.