Posted On: Nov 15, 2023

Today, Amazon EC2 Auto Scaling launches a new capability, instance maintenance policy, for customers to define whether instances are launched before or after existing instances are terminated during instance replacement. This controls how your group replaces instances for operations such as instance refresh, health checks, and rebalancing. From the console, you can choose "policy presets" or provide unique settings that help you optimize for availability or cost.

Until now, the EC2 Auto Scaling service determined the order in which instances were terminated and replaced. Now, with the new instance maintenance policy settings, you can balance SLA commitments and costs during instance replacement operations. Instance maintenance policy settings can be set for all instance replacement operations at the Auto Scaling Group level and overridden for instance refresh operations.

Instance maintenance policy for Amazon EC2 Auto Scaling is generally available in all AWS commercial Regions, including the AWS GovCloud (US) Regions, except China (Beijing, operated by Sinnet), and China (Ningxia, operated by NWCD) Regions.

To learn more, see the EC2 Auto Scaling instance maintenance policy documentation.