AWS regularly performs routine maintenance with minimal disruption to instances. To achieve this we employ a combination of tools and methods across the entire AWS Global infrastructure, such as redundant and concurrently maintainable systems, as well as live system updates. However, in certain cases, we may send scheduled events to customers. This page provides information on scheduled events.
1) What are scheduled events?
Scheduled events are AWS initiated stops, reboots for planned maintenance. Customers are notified in advance regarding Scheduled events via email, EC2 events page and AWS Health Dashboard. Scheduled events are essential for keeping the EC2 fleet healthy and reliable.
2) Why are scheduled events needed?
In some cases, scheduled events may be needed when we detect a hardware degradation. In other cases, scheduled events may be needed to retire older hardware. Lower customer demand for older generation instances combined with higher inefficiency in terms of power and space utilization compared to modern platforms makes them a burden to continue operating and limits our ability to provide newer, more efficient hardware. Lastly, there may be some updates such as BIOS updates for which scheduled events may be needed.
3) What are the best practices to minimize scheduled events?
To minimize scheduled events, we recommend using the newer generation instances. Newer generation instances typically have fewer scheduled events in comparison to older generation instances.
4) What are the types of Scheduled Events? What actions can/should I take for each event?
Scheduled events are typically either reboot events or stop events. In rare cases, you may experience a network maintenance event.
Types of events and actions to be take for each event type:
- Stop event - Stop events are scheduled events that stop the instance during the scheduled time window. When you start the instance again, it is moved to a new host. Alternately, you can stop-start the instance any time prior to the date/time specified in the scheduled event notification to avoid the event. If you use an instance with instance store volumes then you should backup the data prior to the scheduled stop event or before you stop the instance, as the data on the instance store volume is lost when the instance is stopped.
- Reboot event - Reboot events are scheduled events that reboot the instance during the scheduled time window. Instance store data is preserved for a reboot event. You can stop-start the instance any time prior to the date/time specified in the scheduled event notification to avoid the event.
- Network event - Network events are scheduled events where the network connectivity of the instance is lost for a brief period but is restored after the event. The instance continues to run throughout the period. You can stop-start the instance any time prior to the date/time specified in the scheduled event notification to avoid the event.
5) How am I notified regarding scheduled events?
You are notified regarding scheduled events via multiple channels. You are sent an email notification for Scheduled Events. Additionally, the notification is available via the AWS Health Dashboard and EC2 Events page. You can also access event details either using the EC2 Health APIs or via describe-instance-status API. You can also use IMDS to get information on Scheduled events. Lastly, we also generate a CloudWatch event for Scheduled events that can be monitored using Event Bridge rules.
6) How do I know if my instance was sent a scheduled event because it was on degraded hardware or for other reasons?
The email and AWS Health Dashboard notification will specifically detail if the scheduled event is needed because your instance is running on degraded hardware.
7) How do I include instance tags in my scheduled event email notification?
Refer to the documentation to add instance tags in your Scheduled event email notifications.
8) How can I automate my response to scheduled events?
You can monitor CloudWatch events related to scheduled events using EventBridge rules. These events are categorized under the Health service, EC2 service, scheduledChange category in EventBridge rules. Refer to EventBridge rules. You can create a target lambda that stops and starts your instance in response to scheduled events.
9) How can I send scheduled event notifications to additional emails besides the account owner emails?
You can monitor CloudWatch events related to scheduled events using EventBridge rules and define a target SNS Topic that the additional emails can subscribe to. Refer to EventBridge rules.
10) Which Scheduled events can be re-scheduled? How can I reschedule events?
The scheduled event’s email notification will indicate whether if the event can be rescheduled. Typically, if the instance is not on a degraded host, then it can be rescheduled. Refer to Reschedule events.
11) How can I control the timing of Scheduled Events?
EC2 Instance Event Windows lets you define weekly recurring time windows for Scheduled Events that reboot, stop instances or terminate instances. Refer to the Event Windows page.
12) How are scheduled events different from hardware issues that require your instance to be recovered?
Scheduled events are maintenance events for which customers are notified in advance. While some scheduled events pertain to hardware degrades, others are related to hardware retirements and BIOS updates.
In some cases, hardware failures may occur causing an instance to become unreachable and the system status check of the EC2 instance fails. For instances that support automatic recovery, system status check failures automatically trigger recovery. Refer to Recover your instance.