- 최신
- 최다 투표
- 가장 많은 댓글
Auto scaling groups are a great place to start if you need to increase capacity. That said, I would first advise establishing what is causing the high CPU usage. If it is due to high/unexpected application traffic then an auto scaling group is a great idea. However, if you have a rogue process that is utilizing the CPU then an auto scaling group will lead to you creating new EC2 instances and additional costs.
To summarise, the first step needs to be identifying what is causing the high load and plan from there.
Thank you all for your suggestions and answers. Found the issue. AWS cloud watch is starting a new instance every few minutes. I am certain there is some config issue with aws that is triggering it. Will look in that direction for now.
Have you looked into Auto Scaling Group?
You can set up ASG to provide additional server capacity when the load is high and then scale back the amount of servers when the load is reduced.
https://docs.aws.amazon.com/autoscaling/ec2/userguide/what-is-amazon-ec2-auto-scaling.html
Check out this video in Skill Builder: https://explore.skillbuilder.aws/learn/course/external/view/elearning/84/introduction-to-ec2-auto-scaling
The cpu usage is 100% even when it is sitting idle. We are not facing max resource call to scale up.
관련 콘텐츠
AWS 공식업데이트됨 2년 전- AWS 공식업데이트됨 일 년 전
- AWS 공식업데이트됨 2년 전
Capacity is the issue here.
If capacity is the issue then I would recommend implementing auto scaling groups as per the first answer to the question. If CPU is 100% even when idle I would recommend looking at the specific tasks that are using the CPU using task manager. Identifying the specific task causing an issue may help move forward.
Alternatively, what happens if you start a new instance with the same configuration? Does it instantly spike to 100%?
Failing all of that, I'd need a little bit more information on what exactly the Windows servers are used for (web traffic, background processing etc).