Cpu/Memory Low Cloud Watch Alarm for always in alarm state

Question

Hi,
I have configured step scaling policies  for my ECS Service . The cpu & memory utilization always remain under utilization as we are not release  yet. I want the "scale in " cloud watch alarm to go in OK state if the desired number of ECS service is met . How to stop   "scale in "  the cloud watch alarm from staying in an "ALARM" state even though the desired number of instances in the auto scaling group has been met.

Answer

Hi, the alarm's state in CloudWatch only reacts to the metric's value - if the metric is breaching the threshold, the alarm stays in ALARM, even if the scaling group doesn't need to take any new action. There is no option to force the alarm in OK state if the metric keeps breaching the threshold.

Are you concerned that seeing the alarm in ALARM state is giving you a false positive, because the scaling group has reached the configuration you wanted? Do you know you can hide the autoscaling alarms from the alarms list if you don't see value in listing them? I understand this is not your ask, just wanted to share this in case it would help you get a better view of your alarms.

Cpu/Memory Low Cloud Watch Alarm for always in alarm state

相關內容