ApsaraDB RDS for PostgreSQL supports two storage types: ESSD and Premium ESSD. This topic describes the features, differences, and selection guidance for each storage type.
Select a storage type
The following table summarizes when to choose each storage type.
| Scenario | Recommended storage type | Why |
|---|---|---|
| Heavy, sustained I/O loads | Premium ESSD | Use the Buffer Pool Extension (BPE) feature to accelerate I/O performance. |
| Variable I/O patterns with traffic spikes | Premium ESSD | Use the I/O performance burst feature to handle spikes without over-provisioning. |
| Large volumes of infrequently accessed data | Premium ESSD | Use the data archiving feature to move cold data to lower-cost storage. |
| Specific performance tier required (PL1, PL2, or PL3) | ESSD | Choose from three performance levels to match storage performance to your workload. |
ESSD
ESSDs are built on a next-generation distributed block storage architecture. They are integrated with 25 Gigabit Ethernet and Remote Direct Memory Access (RDMA) technologies to reduce one-way latencies and process up to 1 million random read and write requests per second.
ESSDs are available in three performance levels (PLs):
| Performance level | IOPS and throughput | Use case |
|---|---|---|
| PL1 ESSD | Baseline performance level | General-purpose workloads |
| PL2 ESSD | Approximately 2x the IOPS and throughput of PL1 | Workloads that require higher I/O capacity |
| PL3 ESSD | Up to 20x the IOPS and up to 11x the throughput of PL1 | High concurrent requests, high I/O performance, and low read and write latencies |
Premium ESSD
Premium ESSDs are compatible with all features of ESSDs and provide the following additional capabilities:
I/O performance burst: Temporarily exceeds the baseline I/O performance to handle unexpected traffic spikes.
Buffer Pool Extension (BPE): Accelerates I/O performance by using high-speed cache media, particularly for complex queries and temporary data operations.
Data archiving: Moves infrequently accessed data to lower-cost storage to reduce costs.
For more information, see Premium ESSD.
Comparison of storage types
The following table compares the key attributes of ESSD and Premium ESSD.
| Item | ESSD | Premium ESSD |
|---|---|---|
| I/O performance | IOPS varies based on disk specifications and instance types. I/O latency ranges from 100 to 200 microseconds. | Same base performance as ESSD, plus the I/O performance burst and BPE features. |
| Configuration flexibility | You can expand or reduce the storage capacity of an RDS instance. For capacity reduction requirements, see Change instance specifications. | Same as ESSD. |
| Backup method | Snapshot backups | Snapshot backups |
| Time required for backup, read-only instance creation, and instance cloning | A few seconds | A few seconds |
| Storage capacity expansion | Online expansion within a few seconds, with no impact on your workloads. | Online expansion within a few seconds, with no impact on your workloads. |
| Data durability | 99.9999999% (nine nines) | 99.9999999% (nine nines) |
| Supported editions | RDS Basic Edition, RDS High-availability Edition, RDS Cluster Edition | RDS Basic Edition, RDS High-availability Edition, RDS Cluster Edition |
Both ESSD and Premium ESSD use a distributed storage architecture with multiple data replicas to deliver 99.9999999% data durability. For RDS instances that run RDS High-availability Edition or RDS Cluster Edition, a primary/secondary switchover is performed within seconds if the primary instance fails.
For detailed performance specifications, including the maximum capacity per disk, maximum IOPS, and maximum throughput for each storage type, see Block storage performance.
View the storage type of an RDS instance
To check the storage type of your RDS instance:
Log on to the ApsaraDB RDS console.
Go to the Basic Information page of the target RDS instance.
In the Basic Information section, find the storage type.
