PolarDB for PostgreSQL Standard Edition is a cloud-native relational database built on Alibaba Cloud's shared storage architecture. It runs on Elastic Compute Service (ECS) instances backed by enterprise SSDs (ESSDs), delivering high performance and elasticity at a lower cost than the Enterprise Edition.
How Standard Edition differs from Enterprise Edition
Standard Edition and Enterprise Edition share the same core architecture and main features. The difference is in the hardware at the computing and storage layers.
| Layer | Enterprise Edition | Standard Edition |
|---|---|---|
| Computing | Physical machines (no virtualization overhead) | ECS instances |
| Storage | PolarStore PSL4 and PSL5 | Cloud disks: PL0, PL1, PL2, PL3, and AutoPL |
Computing layer: Enterprise Edition compute nodes run on bare physical machines, eliminating virtualization overhead. Standard Edition uses ECS instances.virtualization
Storage layer: Enterprise Edition uses PolarStore (PSL4 and PSL5), where performance metrics depend on compute node specifications and are independent of storage capacity — this gives Enterprise users consistent, predictable I/O regardless of how much data they store.
Standard Edition uses standard cloud disk technology. Performance scales with storage capacity: starting at approximately 50 IOPS per GB for small datasets and increasing as data volume grows, providing a smooth performance curve as your workload expands.
The following figure shows the architecture of PolarDB for PostgreSQL Standard Edition.
Storage options and performance levels
Standard Edition stores data in ESSDs built on a next-generation distributed block storage architecture with 25 Gigabit Ethernet and Remote Direct Memory Access (RDMA). Each ESSD delivers low one-way latency and up to 1 million random read/write IOPS.
ESSDs are available in the following performance levels (PLs):
| Performance level | IOPS | Throughput | Best for |
|---|---|---|---|
| PL0 | Basic | Basic | Development and testing; low-traffic workloads |
| PL1 | 5x PL0 | ~2x PL0 | General-purpose production workloads |
| PL2 | ~2x PL1 | ~2x PL1 | Medium-traffic applications needing higher I/O |
| PL3 | Up to 10x PL2 | Up to 5x PL2 | High-concurrency workloads requiring high I/O and low latency |
| AutoPL | Configurable | Configurable | Variable workloads where I/O demand fluctuates |
Choosing a performance level: Choose a fixed PL (PL1–PL3) when your workload has steady, predictable I/O — select the level based on your throughput requirements. Choose AutoPL when your I/O demand is variable or unpredictable: AutoPL decouples IOPS from storage capacity so you configure IOPS independently, which reduces the overall Total Cost of Ownership (TCO).
For detailed ESSD specifications, see ESSDs.
When an ESSD's storage is exhausted, the disk is locked to read-only operations.
Data reliability
ESSDs use a distributed triplicate mechanism, providing 99.9999999% (nine 9s) data durability. If a hardware failure interrupts service in one zone, data is automatically copied to an available disk in another zone to maintain durability and availability. For details, see Triplicate storage.
Get started with Standard Edition
Purchase: Buy a PolarDB for PostgreSQL Standard Edition cluster on the . If you're new to PolarDB, try the first.
Migrate from ApsaraDB RDS: When purchasing a new cluster, use the migrate data from ApsaraDB RDS feature to migrate at no additional cost.
Migrate from other sources: Use Data Transmission Service (DTS) to migrate from a self-managed PostgreSQL database or an ApsaraDB RDS for PostgreSQL instance. Alternatively, use the pg_dump and pg_restore tools from the PostgreSQL open-source community.
What's next
PolarDB forPostgreSQLdocumentationPolarDB for PostgreSQL free trial pagePolarDB for PostgreSQL official documentationPolarDB for PostgreSQL kernel feature introduction
PolarDB official account (Search for "PolarDB" on WeChat)
Customer support email: polardb-pg-support@list.alibaba-inc.com