PolarDB for PostgreSQL (Compatible with Oracle) supports multi-zone cluster deployment with three high-availability (HA) modes. Each mode offers a different balance of data protection, compute redundancy, and cost. Use this page to compare modes and complete common tasks such as enabling HA, viewing zone configuration, changing the primary zone, and creating a read-only endpoint.
Choose a deployment mode
| Mode | Data distribution | Compute redundancy | Storage cost | Best for |
|---|---|---|---|---|
| Single-zone (hot standby storage cluster disabled) | Primary zone only | Primary zone only | Lowest | Development, testing, small applications |
| Dual-zone (hot standby storage cluster enabled) | Primary + secondary zone, 3 replicas each (6 total) | Primary zone only | Higher | Most production workloads (>80% of use cases) |
| Dual-zone (hot standby storage cluster and compute node enabled) | Primary + secondary zone, 3 replicas each (6 total) | Primary zone + secondary zone | Highest | Large enterprises, high read throughput, mission-critical systems |
To use multi-zone deployment, the target region must contain at least two zones, and the secondary zone must have sufficient computing resources.
How each mode works
Single-zone (hot standby storage cluster disabled)
All database services run in the primary zone. This mode has the lowest cost, but if the primary zone fails, service recovery takes longer than in dual-zone modes. If you need cross-zone high availability, use one of the dual-zone modes instead.
Switching rules: Can be switched to dual-zone (hot standby storage cluster enabled) or dual-zone (hot standby storage cluster and compute node enabled).
Use cases:
Small websites and applications — offload routine operations and maintenance (O&M) to Alibaba Cloud and focus on application development.
Individual users — test and learn PolarDB without a long-term commitment.
Development and testing — take advantage of fast provisioning and flexible scaling to improve R&D efficiency.
Dual-zone (hot standby storage cluster enabled)
Data is distributed across two zones. The primary zone and secondary zone each store a full copy of data with three replicas, for six replicas in total. Compute nodes run in the primary zone. The hot standby storage cluster in the secondary zone serves as the failover target when the primary zone fails, ensuring a high service level agreement (SLA).
We recommend this mode for most production workloads.
Switching rules: Can only be switched to single-zone (hot standby storage cluster disabled).
Billing: Storing data across two zones doubles the number of replicas (6 total), so storage costs are higher than single-zone mode. For billing details, see Billing.
Use cases: Suitable for the majority of production workloads across industries including internet, IoT, online retail, logistics, and gaming.
Dual-zone (hot standby storage cluster and compute node enabled)
Data is distributed across two zones with six replicas in total (three per zone). Both storage and compute are replicated: the secondary zone hosts a hot standby compute node that can serve read traffic. After a failover, the hot standby compute node is promoted to the primary node.
Switching rules: Can only be switched to single-zone (hot standby storage cluster disabled).
Billing:
Compute nodes — hot standby compute nodes in the secondary zone are charged separately. By default, their specifications match those of the primary zone.
Storage — same as dual-zone (hot standby storage cluster enabled): six replicas in total, higher cost than single-zone.
For billing details, see Billing.
To use this mode, contact us.
Use cases: Large and medium enterprises whose production databases handle high read traffic or run intelligent data analysis during peak hours, including financial institutions, online retailers, automobile companies, education providers, and Enterprise Resource Planning (ERP) service providers.
Multi-zone deployment architecture
The following diagram shows the architecture of dual-zone deployment modes.
Enable or switch the HA mode
During cluster creation: Select the HA mode in the Network and Zone section on the purchase page.
Available HA modes depend on the zones and resources in the selected region. Check the purchase page for supported options.
For an existing cluster: In the PolarDB console, go to Clusters, click the cluster ID, and in the Database Distributed Storage section of the Basic Information page, click Switch to HA Mode.
View the zones of a cluster
Log on to the PolarDB console. In the left-side navigation pane, click Clusters. In the upper-left corner of the page, select the region where the cluster resides, then click the cluster ID to go to the Basic Information page.
On the Overview page, view the Zones section.
Change the primary zone
Multi-zone clusters support changing the primary zone, which migrates compute nodes to a different zone. This is useful for disaster recovery drills or when Elastic Compute Service (ECS) instances need to access the cluster from a nearby zone.
For a PolarDB for PostgreSQL (Compatible with Oracle) cluster for which hot standby is enabled, you cannot change the primary zone after you enable tiered storage for cold data for the cluster. To do so, contact us.
Log on to the PolarDB console. In the left-side navigation pane, click Clusters. In the upper-left corner of the page, select the region where the cluster resides, then click the cluster ID to go to the Basic Information page.
On the Basic Information page, click Change Primary Zone.
In the dialog box, configure the Destination Zone and Destination vSwitch parameters, then set Effective Time based on your requirements.
- Destination zone is an existing secondary zone — only compute nodes are migrated; no data migration is needed. Node migration across data centers takes about 5 minutes on average. This is the typical path for disaster recovery drills. - Destination zone is not a secondary zone — data migration is required. Migration time varies by data volume and may take several hours for large datasets. This path is typically used to realign application and database zones for low-latency access.
In the confirmation message, click OK.
After the primary zone changes, the primary endpoint and cluster endpoints remain the same. However, the vSwitch and IP address may change. The operation causes a service interruption of less than 60 seconds.
Create a read-only endpoint for the hot standby compute node
You can create a read-only endpoint for the hot standby compute node when the HA mode is set to dual-zone (hot standby storage cluster and compute node enabled). If you want to use the dual-zone (hot standby storage cluster and compute cluster enabled) deployment, contact us.
The read-only endpoint for the hot standby compute node must be a private endpoint.
Log on to the PolarDB console. In the left-side navigation pane, click Clusters. In the upper-left corner of the page, select the region where the cluster resides, then click the cluster ID to go to the Basic Information page.
In the Database Connections section, click Create Read-only Node Endpoint, then click OK in the message that appears. Use this endpoint to access the hot standby compute node.
