Instance migration due to upgrades at the underlying layer - Elastic Compute Service

This topic describes the events that are related to instance migration due to upgrades at the underlying layer. This topic also describes the recommended migration solutions.

Events

To ensure and improve the stability and performance of Elastic Compute Service (ECS) instances, Alibaba Cloud plans to upgrade and improve the physical infrastructure in specific existing regions and zones. To prevent impacts on existing instances in the regions and zones, Alibaba Cloud generates system events and notifies you by phone calls, text messages, emails, and internal messages. If you receive such notifications, migrate your instances as prompted.

The system events have an event code of SystemUpgrade.Migrate. Perform the following steps to handle a SystemUpgrade.Migrate event:

Check the event notification to identify which instances need to be migrated.
Select a suitable time period to migrate your instances.

Migration solutions and considerations

Note

If you encounter issues when you migrate ECS instances, refer to the FAQ topic in the relevant documentation or contact your account manager.

Instances that reside in virtual private clouds (VPCs) and in the classic network cannot communicate with each other over the internal network. To allow communication between the instances over the internal network, use the recommended migration solutions in the following table to migrate instances during a suitable time period.

Scenario	Recommended migration solution and consideration
The instances reside in the classic network.	Use the following migration solutions to migrate instances to VPCs: Solution 1: Click Migrate Based on Migration Plans. In the Ensure interconnections between the migrated instances and the classic network-type instances specified in the plan over the internal network section, select Yes. Add all ECS instances that require mutual access to the same migration plan. You can specify different migration times to migrate the instances. For more information, see Migrate ECS instances from the classic network to a VPC. Solution 2: Migrate instances based on snapshots and images. After the instances are migrated, you must establish ClassLink connections on the new instances. For more information, see Use OOS to migrate ECS instances and Overview.
The instances reside in VPCs.	Solution 1: Click Clone Based on Snapshots and Images to automatically clone instances by using the ACS-ECS-BulkyCloneInstances public template provided by CloudOps Orchestration Service (OOS). Before you release the original instances, make sure that the new instances run as expected and that traffic is diverted to the new instances. Solution 2: Manually migrate instances in the ECS console. Before you release the original instances, make sure that the new instances run as expected and that traffic is diverted to the new instances. For more information, see Use the snapshot and image features to migrate instance data.
Other cloud resources, such as ApsaraDB RDS instances or ApsaraDB for Redis instances, reside in the classic network.	ECS instances that reside in VPCs cannot directly access RDS instances in the classic network over the internal network. You must change the network type of the RDS instances from the classic network to VPC and enable the hybrid access mode. This way, ECS instances that reside in the classic network or VPCs can connect to the RDS instances over the internal network. For more information, see Best practices for migrating all instances from the classic network to a VPC and Hybrid access to ApsaraDB services.
The instances use images of earlier versions, such as Windows Server 2003 and earlier and CentOS 5.8 and earlier.	After you migrate instances that use images of earlier versions, such as Windows Server 2003 and earlier and CentOS 5.8 and earlier, issues such as kernel panic failures, blue-screen errors, and INACCESSIBLE_BOOT_DEVICE errors may occur. For information about how to resolve the issues, see How do I resolve downtime issues that occur on migrated instances?

Query instance migration events due to upgrades at the underlying layer

Log on to the ECS console.
In the left-side navigation pane, choose Instances & Images > Instances.
In the top navigation bar, select the region and resource group to which the resource belongs.
In the left-side navigation pane, click Events.
View instance migration events due to upgrades at the underlying layer.
If a specific number of events need to be handled, the corresponding number is displayed on the Instance Migration Events Due to Upgrades at Underlying Layer tab, as shown in the following figure.
Select a solution to migrate instances based on your requirements.
For information about how to select a migration solution, see the Migration solutions and considerations section.

What to do next

After the migration, you can retain the public IP addresses of the instances. For more information, see Convert the system-assigned public IP address of an instance in the classic network into an EIP or Convert the static public IP address of an ECS instance in a VPC to an EIP.
After the migration, you can change the private IP addresses of the instances. For more information, see Modify the private IP address of an instance.