What is Power Scheduling?

Power scheduling in the public cloud is the practice of powering-off all hourly-billed resources to as little usage hours as possible. Setting a power schedule to automatically power-off instances at 5pm and back on at 8am on weekdays plus off the entire weekend can reduce compute costs by up to 73%. Customize power schedules to follow standard business hours, or an application's hours of productivity.

In the AWS EC2 (virtual machine) service, instances are mainly charged by: (instance type + operating system + regional pricing) * running hours

‍

Example 1 - Consider a t3.xlarge Linux instance in us-east-1 costs $0.1664/hr

This single instance will cost ~$1,457.66/yr without power scheduling.

Result: After enabling a non-working hours power schedule, the cost is reduced to ~$393.57/yr for an annual savings of $1,064.09 !

‍

Example 2 - Consider a r6i.32xlarge Windows instance in us-west-1 costs $14.848/hr

This single instance will cost ~$130,068.48/yr without power scheduling.

Result: After enabling a non-working hours power schedule, the cost is reduced to ~$35,118.48/yr for an annual savings of ~$94,949.99 !!!

‍

Our team has found it much easier to manage power schedules using E1 compared to configuring schedules via resource tags and having to deploying a large services stack in each account. -E1 Customer

‍

This simple solution can save a huge portion of your cloud bill each month - but why stop there? Easily apply the same logic for other services such as RDS, SageMaker, WorkSpaces, and anywhere resources are billed hourly while powered-on within E1.

What will you do with all that money saved? 🤑