All Products
Search
Document Center

ApsaraDB RDS:Create a disaster recovery instance (set up a disaster recovery instance group)

Last Updated:Mar 30, 2026

Set up a cross-region disaster recovery architecture for ApsaraDB RDS for SQL Server by creating a disaster recovery instance group. A disaster recovery instance group pairs a primary instance with a disaster recovery instance in another region. The disaster recovery instance replicates data from the primary in real time and supports read-only access to offload query load. If the primary instance or its region becomes unavailable, promote the disaster recovery instance to take over — keeping your business running with minimal downtime.

Key concepts

Term Description
Disaster recovery instance group A one-to-one pairing of a primary instance and a disaster recovery instance. For architecture details, see Architecture.
Primary instance The read-write role in the group. Handles all write operations and synchronizes data to the disaster recovery instance in real time.
Disaster recovery instance The secondary role in the group. Serves three purposes: (1) Remote disaster recovery — act as a standby node; promote it to primary if the primary instance fails. (2) Remote read-only access — offload read queries from the primary, enabling local access for multi-region applications. (3) Fast data migration — migrate data to another region by promoting the disaster recovery instance.

Prerequisites

Before you begin, ensure that you have:

Limitations

Primary instance requirements

  • Database version: SQL Server 2017 EE, 2019 EE, or 2022 EE. If your version is not supported, upgrade the major version first.

  • Edition: Cluster Edition

  • Billing method: Subscription or pay-as-you-go. Serverless is not supported.

  • SA permissions: Neither database accounts nor host accounts can have SA permissions.

Supported regions for primary instances:

  • Chinese mainland: China (Qingdao), China (Beijing), China (Shanghai), China (Hangzhou), China (Hohhot), China (Zhangjiakou), China (Shenzhen), China (Guangzhou), China (Ulanqab), China (Heyuan), China (Chengdu)

  • China (Hong Kong)

  • Regions outside China: Thailand (Bangkok), South Korea (Seoul), Philippines (Manila), Japan (Tokyo), Singapore, Malaysia (Kuala Lumpur), Indonesia (Jakarta), US (Silicon Valley), US (Virginia), Germany (Frankfurt), UK (London)

To deploy disaster recovery in regions not listed above, submit a ticket.

Disaster recovery instance requirements

  • Database version: SQL Server EE. The major version must match the primary instance.

  • Edition: Cluster Edition or Basic Edition

  • Billing method: Subscription or pay-as-you-go. Serverless is not supported.

  • Storage capacity: Must be greater than or equal to the primary instance's storage capacity.

  • Collation: Must match the primary instance's collation.

  • Existing instances only: The instance cannot contain any databases. Check the Databases page and delete any existing databases before adding the instance.

Operational limits after the group is created

Once a disaster recovery instance group is active, certain operations are disabled to protect data consistency.

Primary instance restrictions:

  • Cannot create accounts with SA permissions

  • Cannot enable Secure Sockets Layer (SSL) encryption or transparent data encryption (TDE)

  • Cannot change zones, upgrade the major version, or update the minor engine version

  • Cannot modify the character set collation

Disaster recovery instance restrictions:

  • The instance is read-only — you cannot create, modify, or delete databases or accounts

  • Cannot perform backup or restoration (backups run on the primary instance)

  • Cannot create read-only instances

  • Cannot enable SSL encryption or TDE

  • Cannot change zones, upgrade versions, or modify the character set collation

Additional notes

  • A disaster recovery instance group contains exactly one primary instance and one disaster recovery instance.

  • An instance can belong to only one disaster recovery instance group, either as the primary or the disaster recovery role.

  • Databases and accounts created on the primary instance are synchronized to the disaster recovery instance. Privileged accounts on the primary become standard accounts after synchronization.

Billing

Scenario What you pay
Create a new instance as the disaster recovery instance RDS instance fee + DTS data synchronization fee
Add an existing instance as the disaster recovery instance DTS data synchronization fee only (no additional instance fee)

For pricing details, see Billing.

Disaster recovery instance parameters

Regardless of which method you use, configure these parameters when creating or adding a disaster recovery instance:

Parameter Description
Region Select a region different from the primary instance's region. The console shows only supported regions.
Database engine Fixed to the same version as the primary instance. Cannot be changed.
Edition Basic Edition or Cluster Edition. Choose based on your availability needs.
Storage capacity Must be greater than or equal to the primary instance's storage capacity.
Character set collation rule Fixed to match the primary instance's collation. Cannot be changed.

Method 1: Create a new instance as the disaster recovery instance

Use this method when you do not have an existing instance to reuse. The system creates a new RDS for SQL Server instance and automatically configures data synchronization.

There are two entry points to set this up.

Option A: From the primary instance details page

  1. Go to the RDS instance list and click the primary instance ID to open the Basic Information page.

  2. In the Instance distribution section, find Disaster recovery instance and click Add.

    After you click Add, the system automatically creates a disaster recovery instance group on the Global Distributed Cache page. You can go to that page to view the group.

    image

  3. On the disaster recovery instance creation page, configure the parameters listed above. To customize additional settings, see Create an ApsaraDB RDS for SQL Server instance.

  4. Review the cost breakdown on the right — it shows both the RDS configuration fee and the DTS data synchronization fee. Click Confirm order.

  5. After payment, the system redirects you to the disaster recovery instance group details page and begins provisioning:

    • Status changes to Adding secondary instance while the RDS SQL Server instance is created.

    • Status changes to Creating data synchronization task while the DTS synchronization link is set up.

    • When the status is Running, setup is complete.

Option B: From the Global Active Instances page

  1. Create the disaster recovery instance group.

    1. Go to the Global Active Instances page and click Create instance group.

    2. On the Create Global Active Instance Group page, configure the following parameters and click OK.

      Parameter Description

      Instance group name

      A custom name for the group. Start with an uppercase letter, a lowercase letter, or a Chinese character. Can contain digits, underscores (_), and hyphens (-). Length: 2–126 characters.

      Database instance type

      Select SQL Server.

      Instance group type

      Select Disaster Recovery. Suitable for cross-region disaster recovery and disaster recovery drill scenarios. The primary role is read-write; the secondary role is read-only. Supports primary/secondary switchover and promotion.

      Region of primary instance

      Same as the region selected in the RDS console. To use a different region, change the region at the top of the console.

      Primary instance

      Select an RDS for SQL Server instance from the drop-down list. If no instances are available in the current region, click Create primary instance to create one.

  2. Add the disaster recovery instance.

    1. On the Global Distributed Cache page, find the target instance group and click Add secondary instance in the Actions column. image

    2. Configure the parameters listed above. To customize additional settings, see Create an ApsaraDB RDS for SQL Server instance.

    3. Review the cost breakdown on the right and click Confirm order.

    4. After payment, the system redirects you to the disaster recovery instance group details page and begins provisioning:

      • Status changes to Adding secondary instance while the RDS SQL Server instance is created.

      • Status changes to Creating data synchronization task while the DTS synchronization link is set up.

      • When the status is Running, setup is complete.

Method 2: Add an existing instance as the disaster recovery instance

Use this method when you already have an RDS for SQL Server instance that meets the disaster recovery instance requirements. The existing instance must contain no databases before you begin.

  1. Create the disaster recovery instance group. Follow the same steps as Option B, step 1 above.

  2. Add the existing instance to the group.

    1. Go to the RDS instances page. In the top navigation bar, select the region where the disaster recovery instance resides, then click the instance ID.

    2. Click Add to GAD Cluster in the upper-right corner.

    3. In the dialog box, verify the instance information (it will be the secondary role) and select the instance group you created in the previous step. image

    4. Click Check. After all precheck items pass, review the data synchronization fee and click OK. The system redirects you to the disaster recovery instance group details page and creates a one-way DTS data synchronization link:

      • Status changes to Creating data synchronization task while the link is set up.

      • When the status is Running, setup is complete.

What's next