What happens during an EC2 scheduled maintenance event and what preventative actions can I take?
Last updated: 2020-09-18
My Amazon Elastic Compute Cloud (Amazon EC2) instance has a maintenance event scheduled. What happens during a maintenance event? What preemptive actions can I perform before the maintenance window?
Usually Amazon EC2 maintenance is a live-update to minimize customer impact. Sometimes, however, a live-update is not possible, and therefore a scheduled maintenance event is needed.
A scheduled maintenance event might require your instance to undergo one of the following:
The instance actions that your maintenance event requires depends on the instance configuration and the type of maintenance that is being performed.
Depending on your situation, you can perform one of the following actions before the specified maintenance window:
Reschedule the maintenance event
If your instance must undergo a reboot, then you can reschedule the event so that your instance is rebooted at a specific date and time that suits you. For more information, see Rescheduling a scheduled event.
Stop and start your instance
If your instance has an Amazon Elastic Block Store (Amazon EBS) volume as its root device, then complete maintenance by stopping and starting your instance. A stop and start migrates your instance to a new underlying host.
The stop and start changes the status of the scheduled maintenance event to Completed. The actions, when completed, can take up to an hour to be reflected in the status change. For more information, see Viewing scheduled events.
Important: Stopping and restarting the instance changes the public IP address of your instance. It's a best practice to use an Elastic IP address instead of a public IP address when routing external traffic to your instance.
Automatically recover an instance
You can create an Amazon CloudWatch alarm that monitors an Amazon EC2 instance and then automatically recovers the instance if the instance is impaired due to an underlying hardware failure. For more information, see Recover your instance.
Create an instance store-backed AMI
If the instance is based on an instance store or an ephemeral root volume and your instance is scheduled to terminate rather than reboot, you might need to create an instance store-backed AMI, migrate your data, and then relaunch the instance.
Let the maintenance event continue as scheduled
If the maintenance schedule is network connectivity, power maintenance, or a required reboot, you can let the maintenance event continue as scheduled without manual intervention.