- 新しい順
- 投票が多い順
- コメントが多い順
Disabling cores don't effect on instance cost. The reason why you would typically want to disable some of available cores is due to software licensing. If your application needs a lot of memory but can not benefit from extra cores, it can be cost effective to disable some cores to lower licensing cost. Even if you would still be paying for the whole instance capacity.
EC2 pricing [1] is based on the instance capacity. Even if you only use 1 vCPU, all 8 (in your example) would be allocated to you and available for use. There are services like Fargate [2] where pricing is based on vCPU. If your application requires fewer vCPUs and more of another resource, like RAM, then it's recommended to select an instance type that more aligns with your needs. My recommendation is to use the Instance Type feature in the EC2 Console. EC2 Console, on the left menu panel, select Instance Type. In this interface, you can list your requirements (number of vCPUs, amount of RAM, networking, etc) and the interface will display the available instance types including their on-demand costs. From this you can select the best instance type and size for your application or needs.
If later you need to increase (or decrease) the vCPU or RAM, you can change the instance type/size requiring only a start/stop [3].
I also recommend enabling the AWS Compute Optimizer (defaults are free) [4] which can use your running data to recommend the right instance type and size.
[1] https://aws.amazon.com/ec2/pricing/on-demand/
[2] https://aws.amazon.com/fargate/pricing/
[3] https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/ec2-instance-resize.html
関連するコンテンツ
AWS公式更新しました 4ヶ月前
