New Auto Scaling features for controlling individual Amazon EC2 instances

Posted on: Jul 30, 2014

You now have more control over what happens to the Amazon EC2 instances running in your Auto Scaling groups at each stage of their lifecycle, from launch through termination. This makes it easier to update and run software on your EC2 instances, and troubleshoot and fix problems.

Starting today, you can use new instance lifecycle action APIs to hold your newly launched instances in a pending state while you perform actions on them, such as installing software, before they are added to your Auto Scaling group and become in service. Similarly, when instances are terminating, you can temporarily stop the termination to investigate the cause or retrieve logs from an instance before it terminates.

While instances are running in your Auto Scaling group, you can use the new DetachInstances action, and manage them just like instances that are not members of an Auto Scaling group. Combined with the AttachInstances action, you can easily move instances into and out of groups.

You can also place instances into the new Standby state to troubleshoot or perform maintenance on them, then put them back into service once you are finished. When you place an instance in Standby, it will no longer receive new inbound connections from Elastic Load Balancers or count toward your group’s capacity until you put it back into service again.

You can access these new features using the Auto Scaling API and command-line tools. They will be available in the AWS SDKs soon. To learn more about Auto Scaling, please visit the Auto Scaling Developer Guide.