Posted On: Dec 7, 2015

The Amazon Auto Scaling service now allows you to protect instances from termination during a scale-in event.

Amazon Auto Scaling automatically manages your Amazon Elastic Compute Cloud (EC2) capacity. During periods of low utilization, instances are automatically terminated in order to save you money. There may be times, however, when you want to protect certain instances from termination, such as:

  • A worker in an auto-scaled pool that is performing a long-running task
  • An instance that has a special meaning in the auto scaling group, such as the master in a Hadoop cluster

Now you can mark instances as protected from termination. The flag can be set as the instance is launched by Auto Scaling, or dynamically changed throughout the instance’s lifetime.

This feature is free to use. To get started, navigate to the Auto Scaling section of the Amazon EC2 console, select an auto scaling group, go to the instances tab and right-click on an instance to protect it. To learn more about this feature, review the Amazon Auto Scaling documentation.