EC2 Auto Scaling introduces instance lifecycle policy

Today, EC2 Auto Scaling announces a new feature called instance lifecycle policy. Customers can now configure a way to retain their instances when their termination lifecycle hooks fail or timeout, providing greater confidence in managing instances for graceful shutdown.

You can add lifecycle hooks to an Auto Scaling group (ASG) to perform custom actions when an instance enters a wait state. You can choose a target service (e.g., Amazon EventBridge or AWS Lambda) to perform these actions depending on your preferred development approach. Customers use ASG lifecycle hooks to save application state, properly close database connections, back up important data from local storage, delete sensitive data/credentials, or deregister from service discovery before instance termination. Previously, both default results—continue and abandon—led to ASG terminating instances when the lifecycle hook timeout elapsed or if an unexpected failure occurred. With the new instance lifecycle policy, you can now configure retention-triggers to keep your instances in a retained state for manual intervention until you're ready to terminate them again. This policy provides greater confidence in graceful instance termination and is especially helpful for stateful applications running on ASG.

This feature is now available in US East (N. Virginia), US West (Oregon), Europe (Ireland), and Asia Pacific (Singapore). To get started, visit the EC2 Auto Scaling console or refer to our technical documentation.