Disaster recovery solution for ApsaraDB RDS instances
This topic describes the various disaster recovery solutions provided by ApsaraDB RDS to help you improve database availability and data security.
Create disaster recovery instances
RDS SQL Server
RDS SQL Server provides cross-region disaster recovery functionality based on SQL Server Distributed Availability Groups (DAG) and DTS cross-region network connectivity capabilities, offering users a one-stop, cost-effective cross-region disaster recovery solution. The disaster recovery instance group adopts a "primary-secondary" architecture, where the primary role instance supports read and write operations, while the secondary role instance only supports read-only operations. Both primary and secondary role instances must be running Enterprise Cluster Edition (Cluster Edition) 2017 or later. The primary and secondary roles are connected through a one-way data synchronization link, using DAG and DTS to achieve real-time cross-region data synchronization, ensuring data consistency while providing nearby access experience. In disaster recovery scenarios, you can quickly restore application access through the primary/secondary switchover or disaster recovery promotion features of the disaster recovery instance group.
Scenarios: Suitable for cross-region disaster recovery, disaster recovery drills, or other scenarios requiring disaster recovery capabilities.
Deploy multi-zone instances
Alibaba Cloud provides efficient and reliable cloud computing services for multiple regions worldwide, with each region containing multiple zones. Zones within the same region feature low network latency and fault isolation, ensuring high availability and disaster recovery capabilities for cross-zone deployments.
RDS single-zone instances run on two physical servers within the same zone, providing higher availability than traditional physical servers. This architecture has the following features:
Infrastructure redundancy: Single-zone racks, air conditioners, circuits, and network equipment all adopt redundant designs to minimize the impact of hardware failures on services.
Data replication mechanism: Supports asynchronous or semi-synchronous data replication methods to ensure data consistency between primary and secondary nodes.
Efficient primary/secondary switchover: When the primary node fails, the system can quickly switch to the standby node to ensure service continuity.
To further improve database availability and disaster recovery capabilities, ApsaraDB RDS provides multi-zone instance deployment solutions. This solution deploys physical servers in different zones. When one zone fails, traffic can be switched to another zone within a short period of time, significantly enhancing the system's risk resistance. The entire switchover process is transparent to users and does not require changes to the application code.
Method: Migrate an ApsaraDB RDS for SQL Server instance across zones
When a primary/secondary switchover occurs, the connection to the RDS instance is closed and the application needs to reconnect to the RDS instance.
DTS-based disaster recovery
You can use Alibaba Cloud Data Transmission Service (DTS) to synchronize data from a database in your self-managed data center or a self-managed database on your Elastic Compute Service (ECS) instance to your RDS instance in any region in real time. If unexpected exceptions, such as data center damages, occur, you can obtain database data from your RDS instance. For more information, see Overview of DTS data synchronization solutions.