EC2 Auto Scaling now lets you set a default instance warm-up time for all instance scaling and replacement actions
You can now set a default instance warm-up time for all scaling activities, health check replacements, and other replacement events in the Auto Scaling instance lifecycle. Amazon EC2 Auto Scaling is a service that allows you to automatically scale and manage logical groups of instances, known as Auto Scaling groups, that serve your application. EC2 Auto Scaling does this by monitoring various metrics, such as CPU utilization and application demand, to determine if an instance needs to be replaced, removed from, or added to your Auto Scaling group. Setting the default instance warm-up time parameter can simplify your Auto Scaling group configuration by ensuring that any scaling and replacement policies are aware of the time your instances typically take to be ready to serve demand.
You can now set a default instance warm-up time for all scaling activities, health check replacements, and other replacement events in the Auto Scaling instance lifecycle. Amazon EC2 Auto Scaling is a service that allows you to automatically scale and manage logical groups of instances, known as Auto Scaling groups, that serve your application. EC2 Auto Scaling does this by monitoring various metrics, such as CPU utilization and application demand, to determine if an instance needs to be replaced, removed from, or added to your Auto Scaling group. Setting the default instance warm-up time parameter can simplify your Auto Scaling group configuration by ensuring that any scaling and replacement policies are aware of the time your instances typically take to be ready to serve demand.
Previously, you could only set a warm-up time for select scaling and replacement events — instance refresh, target tracking, and step scaling policies — and these needed to be configured individually. Now, by setting a default warm-up time parameter that applies to the whole Auto Scaling group, you can easily ensure that all instance scaling and replacement events — instance refresh, manual and dynamic scaling policies, scheduled actions, and health check replacements — use the same warm-up time to aggregate metrics to Amazon CloudWatch and determine your group capacity. Specifically, if your instances tend to spend a known amount of time to get ready to serve traffic after they are launched (e.g., to pre-load application data), setting the default instance warm-up time can help you avoid inadvertent scaling or replacement events in the following ways. First, instances won’t be counted toward your Auto Scaling group’s desired capacity until the warm-up time elapses. Second, instances’ CloudWatch metrics won’t be used to start scaling or replacement actions until after the warm-up time elapses.
You can set the default instance warm-up time on your EC2 Auto Scaling group via the DefaultInstanceWarmup parameter through the AWS Management Console, the AWS SDKs, the AWS Command Line Interface (CLI), and CloudFormation. This feature is available in all public AWS Regions. For information on this feature, visit this page.