If a host on which your Elastic Compute Service (ECS) instance resides has potential risks or fails, Alibaba Cloud sends a system event to notify you of the situation. You can handle the event. When you handle the event, the system takes the default maintenance action to restart the instance. If you do not want the instance to be automatically restarted, you can change the maintenance action for the instance by modifying its maintenance attribute.
Background information
Note You can modify instance maintenance attributes but the new maintenance attributes
do not affect ongoing operations. For example, if you modify the maintenance attribute
of an instance that is being automatically restarted, the new maintenance attribute
cannot stop the restart operation or change it into a different operation.
Instance maintenance attribute | System event | Supported instance | Description |
---|---|---|---|
Automatically Restart (Default) |
|
All instances that support system events. | Restores the instance to the state that it was in before the O&M task related to a
system event is executed.
|
Stop |
|
All instances that support system events. | Puts the instance into the Stopped state. This maintenance attribute is applicable to scenarios in which failovers or switchovers are performed at the application layer for disaster recovery purposes to prevent conflicts between multiple nodes in service. |
Automatically Re-deploy |
|
Instances that depend on host hardware, such as instances that have local disks attached
or that support Software Guard Extensions (SGX) encrypted computing. For information
about the instance families of such instances, see Instance family.
Note After an instance is redeployed, the data on its local disks is deleted and SGX is
reset.
|
Redeploys the instance to another host to continue providing services. |
To make better use of instance maintenance attributes, we recommend that you use multiple
methods to improve the operation error tolerance of your business and reduce the impacts
of O&M. Examples:
- Configure core applications such as SAP HANA to automatically start on system startup. This helps prevent business interruptions.
- Enable the automatic reconnection feature for your applications. For example, allow applications to automatically connect to MySQL, SQL Server, or Apache Tomcat.
- Deploy multiple ECS instances in a cluster if you use Server Load Balancer (SLB). While an ECS instance is being automatically restarted, other ECS instances can continue to provide access to your services.
- Back up data on local disks on a regular basis to ensure that redundant data copies are available for you to redeploy instances.