Async replication is a feature that protects data across regions or across zones within the same region based on the data replication capability of Elastic Block Storage (EBS). This feature allows data to be asynchronously replicated between disks across regions or across zones within the same region for disaster recovery. You can use this feature to implement disaster recovery for critical business to protect data in your databases and improve business continuity.
Use scenarios
The async replication feature is suitable for users that require high data security
and need to implement disaster recovery and cross-region migration of business data.
- Disaster recovery scenario
If the primary disk in a replication pair fails, you can use the failover feature to switch the primary and secondary disks over. During this switchover, the failover feature disconnects the original replication link and fails services over to the disaster recovery system by attaching the new primary disk (original secondary disk) to an Elastic Compute Service (ECS) instance that is used for disaster recovery.
- Cross-region migration scenario
If you want to migrate your business data across regions, you can use the reverse replication sub-feature, instead of the image or snapshot replication feature.
Feature description
Note The async replication feature is supported in the following regions and zones:
- China (Hangzhou): Zone H and Zone I
- China (Shanghai): Zone B, Zone E, and Zone F
- China (Beijing): Zone F, Zone G, and Zone J
- China (Shenzhen): Zone D
- China (Heyuan): Zone A and Zone B
- China (Chengdu): Zone A and Zone B
- China (Hong Kong): Zone B and Zone C
- Singapore (Singapore): Zone B and Zone C
- US (Silicon Valley): Zone A and Zone B
- US (Virginia): Zone A and Zone B
How the async replication feature works
The async replication feature allows data to be asynchronously replicated from an
enhanced SSD (ESSD) (primary disk) to another ESSD (secondary disk) that has the same
specifications. The primary and secondary disks can reside in different zones within
the same region or reside in different regions. If the primary disk fails, you can
fail over to the secondary disk. After the primary disk recovers, you can restore
data from the secondary disk. The following figures show how the async replication
feature works when the primary and secondary disks reside in different zones within
the same region or reside in different regions.
- The primary and secondary disks reside in different regions
- The primary and secondary disks reside in different zones within the same region
Operations
The following table describes the operations supported by the async replication feature.
| Operation | Description | References |
|---|---|---|
| Create a replication pair | Before you can use the async replication feature to implement disaster recovery across regions or across zones within the same region, you must create a replication pair. | Create a replication pair |
| Enable the async replication feature | After you create a replication pair, you must enable the async replication feature for the pair to replicate data from the primary disk to the secondary disk cross regions or across zones within the same region on a periodic basis. | Enable async replication for replication pairs |
| Implement disaster recovery | After you create and activate a replication pair, if the primary disk fails, you can use the failover and reverse replication sub-features of async replication to implement disaster recovery. | Implement disaster recovery |
| Delete a replication pair | The functionality of disks in a replication pair is limited. If an existing replication pair is no longer needed to implement disaster recovery across regions or across zones or needs to be replaced, you can delete it. | Delete a replication pair |
Terms
The following table describes the terms related to the async replication feature.
| Term | Description |
|---|---|
| asynchronous replication | The async replication feature supports only asynchronous replication. Unlike synchronous replication, asynchronous replication replicates data from a disk to another disk across regions or across zones within the same region on a periodic basis. In asynchronous replication, the data on the source disk may be inconsistent with that on the destination disk. |
| primary disk | The disk from which to replicate data for disaster recovery. The primary disk is also called the source disk. |
| secondary disk | The disk to which to replicate data. The secondary disk is also called the destination disk. |
| recovery point objective (RPO) | The amount of data that may be lost due to a disk exception. RPO is measured by time. It is used as a data metric and holds a default value of 15 minutes in async replication. This value indicates that data written to a primary disk from up to 15 minutes prior may be lost when an exception occurs on the disk. |
| recovery time objective (RTO) | The duration of time that it takes for a primary disk to recover after an exception occurs on the disk. RTO is used as a data metric in async replication. For example, if the value of RTO is 1 hour, a primary disk can have its data restored and be back in service within 1 hour after an exception occurs on the primary disk. |
| replication pair | The replication relationship that is established between a primary disk, a secondary disk, and configurations for asynchronous replication. |
| failover | A sub-feature of async replication that allows you to enable read and write permissions on the secondary disk and fail over to the secondary disk if the primary disk fails. |
| reverse replication | A sub-feature of async replication that can reverse replication relationships to replicate data from secondary disks to primary disks. |
Billing
You are charged for the use of async replication on a subscription basis and must pay bandwidth fees.
Precautions
The following table describes the limits that apply to the async replication feature.
| Item | Limits |
|---|---|
| Replication pairs that can be created per zone in an Alibaba Cloud account | 16 |
| Replication pairs that can be created for a single disk | 1 |
| Replication cycle | 15 minutes (Data is asynchronously replicated from a primary disk to a secondary disk every 15 minutes.) |
| Primary disk category | A primary disk must be an ESSD used as a data disk and uses the pay-as-you-go billing method. |
| Secondary disk category | A secondary disk must be of the same disk category and have the same performance level and capacity as the associated primary disk. |
The limits described in the following table apply to primary and secondary disks when
you use the async replication feature.
Note:
| Item | Support by the primary disk | Support by the secondary disk |
|---|---|---|
| Read and write operations | √ | ×① |
| Disk deletion | × | × |
| Disk initialization | × | × |
| Disk resizing | × | × |
| Disk attaching | √ | × |
| Snapshot creation | √ | √② |
| Rollback based on snapshots | √ | × |
| Disk category change | × | × |
| Performance level change | × | × |
| Disk encryption | × | × |
| Multi-attach | × | × |
| Disk migration along with instances | × | × |
- ①: After a replication pair is activated, the secondary disk enters the read-only state and no users have write permissions on the disk.
- ②: Due to RPO, data of a snapshot created for a primary disk may not be consistent with that of a snapshot created at the same time for the associated secondary disk.