Replication pairs enable continuous, asynchronous data replication between two ESSD disks for disaster recovery. Data on the primary disk is automatically replicated to the secondary disk every 15 minutes, protecting against zone or regional failures.
Quick decision guide
Use case |
Goal |
Recommended approach |
Protect a single disk within a region |
Near-zero data loss with local redundancy |
Create a cross-zone replication pair |
Protect a single disk across regions |
Regional disaster protection |
Create a cross-region replication pair |
Protect multiple related disks |
Application-consistent failover |
How it works
A replication pair consists of:
Primary disk: The source disk containing your production data
Secondary disk: The destination disk that receives replicated data
The disks must be in different zones—either within the same region (cross-zone replication) or in different regions (cross-region replication).
After activation, the system performs an initial full synchronization, then continuously replicates changes at regular intervals. The default Recovery Point Objective (RPO) is 15 minutes. Enable Replication Time Control (RTC) to reduce the RPO to 10 minutes or less. In a disaster scenario, fail over to the secondary disk with minimal data loss.
Billing
Replication pairs use pay-as-you-go billing. Charges include:
Replication pair service (based on disk capacity)
Snapshot storage (for backup snapshots)
RTC bandwidth (if enabled for cross-region replication)
Enabling RTC for cross-zone replication within the same region incurs no additional charge. Cross-region replication with RTC enabled incurs additional bandwidth charges.
For pricing details, see Block Storage pricing.
Prerequisites
Before you begin, ensure that you have:
Two ESSD or ESSD AutoPL disks with identical capacity and performance level
The disks in different zones (same region or different regions)
The secondary disk in Unattached state
Neither disk undergoing resize or modification
(Recommended) A snapshot of the secondary disk to back up existing data
When you activate the replication pair, data on the secondary disk is overwritten by data from the primary disk. This operation is irreversible. Create a snapshot of the secondary disk before activation to preserve existing data.
Create a replication pair
-
Log on to the EBS console.
NoteFirst-time users must create the required service-linked role when prompted. See Service-linked role for EBS.
In the left navigation pane, choose Enterprise-level Features > Async Replication.
In the top navigation bar, select the region of your primary disk.
On the Primary Site tab, click Create Replication Pair.
-
Configure the following parameters:
Parameter
Description
Billing Method
Pay-as-you-go (only supported method)
Region and Zone
Region and zone of the primary disk
Disk
ID of the primary disk
Disaster Recovery Region and Zone
Destination zone. For cross-zone DR: same region, different zone. For cross-region DR: different region.
Destination Disk
ID of the secondary disk. Must match the primary disk's capacity and performance level.
Enable Replication Time Control (RTC)
Optional. Select Yes to guarantee RPO ≤ 10 minutes using high-quality bandwidth. Cross-zone replication within the same region incurs no additional charge. Cross-region replication incurs additional charges. See Async replication overview.
Risk Disclosure
Read the risk disclosure and select I am aware of the risks to acknowledge that data on the secondary disk will be overwritten.
Replication Pair Name
Identifier for the replication pair
Description
(Optional) Description for reference
Complete the payment process.
Activate the replication pair
After creating the replication pair, activate it to start data replication.
In the replication pair list, locate your replication pair and click Activate in the Actions column.
-
In the Activate Replication Pair dialog box:
-
(Recommended) Click Create Snapshot to back up the secondary disk.
ImportantActivation overwrites all data on the secondary disk. Create a snapshot now if you haven't already backed up the data.
-
Choose an activation method:
Copy Data (one-time sync): Performs a single replication, then sets status to Stopped. Use this for testing or manual control scenarios.
Confirm (automatic sync): Starts continuous replication with 15-minute RPO (or ≤10 minutes if RTC is enabled). Use this for production disaster recovery.
-
-
Monitor the replication status. After activation, the status changes to:
Syncing: Initial synchronization in progress
Normal: Replication pair is active and syncing continuously
Stopped: Replication is paused
What's next
Perform a disaster recovery drill: Test your DR setup to verify it works as expected.
Delete a replication pair: Remove replication pairs that are no longer needed.
Create a replication pair-consistent group: Protect multiple related disks with application-consistent failover.
FAQ
Why can't I select a disk in the destination zone?
The dropdown only shows disks that meet all requirements: ESSD or ESSD AutoPL type, same capacity and performance level as the primary disk, and unattached state.
If you don't see any disks, create a new ESSD disk in the destination zone with matching specifications and keep it unattached.
What happens to the secondary disk during replication?
The secondary disk becomes read-only during active replication. You cannot attach it to an ECS instance or write data to it directly. To use the secondary disk, stop the replication pair first.