For each scaling group, you can control the time to add instances (scale-out) to it or delete instances (scale-in). You can scale the scaling group manually by adding or removing instances, or can allow AS to execute this process automatically by using a scaling policy.
When the scaling group is scaled in automatically, it needs to know which instances should be terminated first, and the selection is based on the removal strategy.
During scale-in, you can prevent specified instances from being terminated by AS by using instance protection.
AS will determine which CVM should be removed based on the removal policy during scale-in. You can choose from the following two removal policies:
Note: No matter whether it is the newest or oldest servers to be deleted, AS will delete the automatically created CVMs first, and then delete manually added CVMs.
There are two ways to set this up:
Was this page helpful?