To ensure and improve the performance and stability of PolarDB-X 1.0 (formerly DRDS), the physical infrastructure in some regions and zones may go through upgrades. This topic provides a migration solution to help you evaluate and migrate the instances affected by such upgrades.
Limitations
This migration solution is valid only for dedicated instances. If you use shared instances, contact us to migrate the shared instance to a dedicated instance.
The Elastic Compute Service (ECS) instances of the classic network type are discontinued. Therefore, when you migrate a PolarDB-X 1.0 instances to other target availability zones and switch to instances of the VPC type.
Description of instance types
PolarDB-X 1.0 instances are divided into three types. Migration methods vary based on the network types of your PolarDB-X 1.0 instances. To view the basic information of an instance, perform step 1 to step 4 in the procedure for the corresponding migration method.
Classic network type
The Basic Information page of an instance of the classic network type, the node type and the network type are classic network, and no VPC endpoint is displayed in the Connection Information section.
Classic network and VPC hybrid type
The node type is classic network. The network type is VPC and the VPC endpoint is displayed in the Connection Information section.
VPC type
The node type and the network type are VPC, and the VPC endpoint is displayed in the Connection Information section.
Migrate an instance of the classic network type to another zone
Limitations
DDL statements, such as CREATE, and GRANT USER statements cannot be executed during the migration.
Service disruptions that last minutes may occur during the migration. We recommend that you perform the migration during off-peak hours.
Usage notes
You cannot migrate a shared instance directly. Contact us to migrate the shared instance to a dedicated instance first.
If the version of your PolarDB-X 1.0 instance is 5.2 or earlier, you need to upgrade the instance to the latest 5.2 version. Instances of earlier versions do not support VPC.
All ApsaraDB RDS instances that are attached to the PolarDB-X 1.0 instance must be changed to the VPC type. For more information, see Change the network type of an ApsaraDB RDS instance. After you change the type of an RDS instance from classic network to VPC, log on to the PolarDB-X 1.0 console and fix the database shard connections. For more information, see Fix database shard connections.
Procedure
Log on to the PolarDB-X console.
In the top navigation bar, select the region where your instance is deployed.
In the left-side navigation pane, click Instance List.
Find the instance that you want to manage and click the instance ID to go to the Basic Information page.
In the Common Actions section, click Switch to VPC.
Select the VPC and vSwitch in the dialog box that appears. If you want to retain the classic network endpoint, select Retain Original Classic Network Endpoint and click OK.
A VPC endpoint is created for the PolarDB-X 1.0 instance. The original classic network endpoint is temporarily retained. Migrate your ECS instance to the VPC at your earliest opportunity and use the VPC endpoint to access the PolarDB-X 1.0 instance. Otherwise, you cannot access the database after the classic network address is released.
After the instance type is changed, click Migrate nodes to a VPC in the Common Actions section.
Configure the parameters and click OK.
Migrate an instance of the classic network and VPC hybrid type to another zone
Before you begin
If the version of your PolarDB-X 1.0 instance is 5.2 or earlier, you need to upgrade the instance to the latest 5.2 version. Instances of earlier versions do not support VPC.
All RDS instances that are attached to the PolarDB-X 1.0 instance must be changed to the VPC type. After you change the type of an RDS instance from classic network to VPC, log on to the PolarDB-X 1.0 console and fix the database shard connections. For more information, see Fix database shard connections.
Procedure
Log on to the PolarDB-X console.
In the top navigation bar, select the region where your instance is deployed.
In the left-side navigation pane, click Instance List.
Find the instance that you want to manage and click the instance ID to go to the Basic Information page.
Click Migrate nodes to a VPC in the Common Actions section.
Configure the parameters and click OK.
Migrate an instance of the VPC type to another zone
Limitations
The migration may take one to two minutes. We recommend that you migrate your instance during off-peak hours to minimize business impact.
Service disruptions that last seconds may occur during the migration. Make sure that your application client supports automatic reconnection.
Make sure that customer application has access to the VPC.
Procedure
Log on to the PolarDB-X console.
In the top navigation bar, select the region where your instance is deployed.
In the left-side navigation pane, click Instance List.
Find the instance that you want to manage and click the instance ID to go to the Basic Information page.
In the Common Actions section, click Switch Zone.
In the dialog box that appears, select the Target Zone from the drop-down list and click OK.