Create and start a replication pair-Elastic Compute Service(ECS)-阿里云帮助中心

To enable disaster recovery for a disk, you can create and start a replication pair to asynchronously replicate data from a primary disk to a secondary disk in a different region or zone.

Prerequisites

The primary disk and secondary disk must meet the following requirements:

Only ESSD or ESSD AutoPL disks are supported. The disks must not be undergoing a resize or specification change.
The secondary disk must have the same disk type, performance level, and capacity as the primary disk.
The primary disk and secondary disk cannot be in the same zone.
The secondary disk must be in the Unattached state.

Asynchronous replication copies data from the primary disk to the secondary disk. To prevent replication failure, the secondary disk must be in the Unattached state during this process.
(Strongly recommended) Create a manual snapshot of the disk to back up its data.

Important
Asynchronous replication replicates data from the primary disk to the secondary disk, overwriting all original data on the secondary disk. This operation is irreversible. We strongly recommend that you back up your disk data. Creating a snapshot incurs charges. For more information, see Snapshot pricing.

Procedure

Log on to the EBS console.

When you log on to the EBS console for the first time, create a service-linked role for EBS as prompted. For more information, see Service-linked role for EBS.
In the left navigation bar, select Enterprise-level Features > Asynchronous Block Storage Replication.
In the upper-left corner of the page, select a region and resource group.

Create a replication pair.

On the Primary Site or Secondary Site tab, click Create Replication Relationship.

In the right-side panel that appears, configure the parameters and follow the on-screen instructions to create the replication pair.

The following table describes the key parameters.

Parameter	Description
Region and Zone	The region and zone of the primary disk.
Cloud Disk	The ID of the primary disk.
Service-linked Role	The default value is Created.
Disaster recovery region and zone	Select the region and zone of the secondary disk. For cross-zone disaster recovery, select a different zone in the same region as the primary disk. For cross-region disaster recovery, select a region different from that of the primary disk.
Destination Disk	Select the ID of the secondary disk. The secondary disk must have the same specifications and capacity as the primary disk.
Risk Disclosure & Compliance Commitment	Read and select I am aware of the risks and agree to the following compliance commitments.
Enable replication time control (RTC)	If you select Yes, an RPO <= 10 minutes SLA guarantee is provided for data replication from the primary disk to the secondary disk. You can modify this setting after creation. For more information, see Data replication time control (RTC).

Start the replication pair.
1. In the replication pair list, find the pair that you created and click Start in the Actions column.
2. In the Activate Replication Pair dialog box, start the replication pair.
  1. Click Create Snapshot. Because asynchronous replication overwrites existing data on the secondary disk, we recommend creating a snapshot to back it up first.
  2. Choose a synchronization mode based on your requirements.
    - One-time synchronization
      
      Click Copy Now to immediately start a one-time data replication. After the replication is complete, the replication pair enters the Stopped state. You can click Copy Now again to perform another one-time synchronization.
      
      Asynchronous replication must be in the Created, One-time Syncing, or Stopped state.
    - Automatic synchronization
      
      Click OK. The system automatically and continuously replicates data based on the RPO.
      
      Asynchronous replication must be in the Created, Syncing, Normal, or Stopped state.
      
      When the status of the replication pair changes to Normal, the initial data synchronization is complete. The system then automatically replicates data every 15 minutes.