We have AWS auto-scaling policy configured for our Linux machine, but for some reason when I checked the scaling activity I realised that it triggers multiple 2-3 machines during very short span of minutes as you can see here:

Activity History

Following is my auto-scaling policy: Scaling Policies

Following is my auto-scaling group description:

Desired: 1
Min: 1
Max: 5
Default Cooldown: 300
Health Check Type: EC2
Health Check Grace Period: 300

is it the health check grace period 300 or default cool down time? why when there is a load on CPU of more than 30% for 300 seconds it tries to load multiple machines instead of 1

1 Answer 1


It looks like a fresh EC2 instance takes about 15 to 20 minutes to initialize and start working, but your cooldown is only 5 minutes. So 5 minutes after increasing your desired count, your new EC2 instance is still initializing, but the CPU is checked again by Auto Scaling and it's still high, so Auto Scaling increases the desired count again.

Try increasing your cooldown to 900 (15 minutes) or even 1200 (20 minutes).

  • I have made these changes, let me watch it for few weeks.
    – T_Flows
    Feb 4, 2017 at 15:42

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .