Help Center> Data Replication Service> Real-Time Synchronization> Task Management> Task Life Cycle> Performing a Primary/Standby Switchover
Updated on 2023-04-26 GMT+08:00
View PDF

Performing a Primary/Standby Switchover

When you create a synchronization task, you can select the DRS instance type. The DRS instance can be a single node instance or a primary/standby instance. The primary/standby deployment can help you improve task reliability. If the task on the primary node fails, the standby node will take over the task and continue the synchronization.

You can select the DRS instance type in the following scenarios:

  • To the cloud
    • MySQL->MySQL
    • MySQL->GaussDB(for MySQL)
    • MySQL->GaussDB(DWS)
  • From the cloud
    • MySQL->Kafka
    • GaussDB(for MySQL)->Kafka
    • GaussDB primary/standby -> Kafka
  • Self-built -> Self-built
    • MySQL->Kafka
    • Oracle->Kafka
    • GaussDB primary/standby -> Kafka

Prerequisites

  • You have logged in to the DRS console.

Scenarios

  1. On the Data Synchronization Management page, click Create Synchronization Task.
  2. On the Create Synchronization Instance page, configure the task name, description, and the synchronization instance details, and click Next.

    Figure 1 Synchronization instance information

  3. Return to the Data Synchronization Management page, you can find that there are two subtasks displayed under the synchronization task you created.

    Figure 2 Primary/Standby tasks

  4. After a synchronization task is configured and started, DRS will start the task on the primary node, and the task on the standby node is suspended.

    Figure 3 Before the primary/standby switchover

  5. If the task on the primary node is abnormal, DRS automatically starts the task on the standby node to continue the synchronization.

    Figure 4 After the primary/standby switchover

Parent topic: Task Life Cycle