This topic introduces the life cycle management of ECS instances.

There are two types of ECS instances that are added to scaling groups: automatically created and manually added instances.

Automatically created ECS instances

Automatically created ECS instances are automatically created according to scaling configuration and rules.

Auto Scaling manages the lifecycle of this ECS instance type. It creates ECS instances during scale-up and stops and releases them during scale-down.

Manually added ECS instances

Manually added ECS instances are manually attached to a scaling group.

Auto Scaling does not manage the lifecycle of this ECS instance type. When this ECS instance is removed from a scaling group, either manually or as the result of a scaling-down activity, Auto Scaling does not stop or release the instance.

Instance status

During its lifecycle, an ECS instance is in one of the following states: Pending:

  • The ECS instance is being added to a scaling group. For example, Auto Scaling creates the ECS instance and adds it to the Server Load Balancer instance, or to the RDS access whitelist.
  • InService: The ECS instance has been added to a scaling group and is functioning correctly.
  • Removing: The ECS instance is being removed from a scaling group.

Instance health status

An ECS instance may be in the following health conditions:

  • Healthy
  • Unhealthy

ECS instances are regarded as unhealthy when they are not Running(Running). Auto Scaling automatically removes unhealthy ECS instances from scaling groups.

  • Auto Scaling only stops and releases automatically created unhealthy instances.
  • It does not stop and release manually added ones.