When AWS detects irreparable failure of the underlying host for your instance, it schedules the instance to either stop or terminate, depending on the instance's root volume type.
-
If the instance has an Amazon EBS root volume, the instance is scheduled to stop.
-
If the instance has an instance store root volume, the instance is scheduled to terminate.
For more information, see Instance retirement.
Important
Any data stored on instance store volumes is lost when an instance is stopped, hibernated, or terminated. This includes instance store volumes that are attached to an instance that has an EBS volume as the root device. Be sure to save data from your instance store volumes that you might need later before the instance is stopped, hibernated, or terminated.
Actions you can take
Actions you can take for instances with an EBS root volume
When you receive a scheduled instance-stop event notification,
you can take one of the following actions:
-
Wait for scheduled stop: You can wait for the instance to stop within its scheduled maintenance window.
-
Perform manual stop and start: You can stop and start the instance yourself at a time that suits you, which migrates it to a new host. This is not the same as rebooting the instance. For more information, see Stop and start Amazon EC2 instances.
-
Automate stop and start: You can automate an immediate stop and start in response to a scheduled
instance-stopevent. For more information, see Running operations on EC2 instances automatically in response to events in AWS Health in the AWS Health User Guide.
Actions you can take for instances with an instance store root volume
When you receive a scheduled system-retirement event
notification, and you want to retain your data, you can take the following
actions:
-
Launch a replacement instance from your most recent AMI.
-
Migrate all necessary data to the replacement instance before the instance is scheduled to terminate.
-
Terminate the original instance, or wait for it to terminate as scheduled.
For more information about the actions you can take, see Instance retirement.
