iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Yes—AWS staging resources can be scheduled to turn off outside working hours without changing application code, if the workload is safe to stop and the right IAM permissions are configured. In one published case study, Rajeev reports using four EventBridge Scheduler schedules to start and stop an EC2 instance and scale an ECS Fargate service down overnight. The author estimated a 87.5% reduction in staging compute costs, from about $19.18 to $2.40 per month; those are estimates for that setup, not independently audited results or a forecast for other AWS accounts.
What the reported 87% saving actually covers
Rajeev’s Aug. 24, 2026, DEV Community case study describes a staging setup with one EC2 instance and one ECS Fargate service. The team says it actively used staging for about three hours a day and scheduled the resources to run from 9 p.m. to midnight Indian Standard Time (IST). The article reports estimated monthly compute costs of about $19.18 before scheduling and $2.40 afterward: a stated saving of $16.78 per month, or 87.5%. It also gives an estimated annual saving of $201.35. Read the case study.
The percentage is consistent with running the scheduled compute for three hours rather than 24, but it is not an AWS benchmark and does not establish that every charge stopped. The figures are the author’s estimates for that example; the article does not provide an independently verified bill. Your savings will depend on actual usage, resource configuration, Region, and services that remain allocated.
How the four schedules worked
EventBridge Scheduler invokes AWS API actions at specified times. In the reported arrangement, a dedicated IAM execution role authorized those actions; EC2 and ECS then handled their respective requests. The setup used no Lambda function and, according to the author, made no application-code changes. It still required AWS configuration: schedules, target actions, and permissions.
#1 Best Overall
| Schedule | Target action in the case study | Intended effect |
|---|---|---|
| EC2 start | Start the instance | Make the staging instance available for the scheduled work window. |
| EC2 stop | Stop the instance | Stop the instance outside that window. |
| ECS scale up | Call ECS UpdateService with desired count set to one | Run one task for the staging service. |
| ECS scale down | Call ECS UpdateService with desired count set to zero | Stop the service’s desired task count outside the window. |
The example’s schedules used 9 p.m. and midnight IST. That timing is specific to the author’s team; choose your schedule from observed access and workload needs, not from the case study’s hours.
Check operational safety before scheduling shutdowns
A staging environment can have work to do even when nobody is actively testing it. Before turning resources off, establish which systems depend on them and when those dependencies run.
Rank #2
- Review access patterns and workload logs to determine when staging is actually used.
- Identify overnight jobs, integrations, test runs, deployments, and dependencies that require the environment to stay available.
- Allow for startup and readiness time before people or automated checks need the service.
- Document a manual override and make sure the people responsible for staging know how to use it.
A scheduled shutdown is a good fit only while the usage pattern remains predictable. AWS Well-Architected guidance recommends supplying resources dynamically to match demand and notes that scheduled scaling can suit consistent patterns. If demand changes, revisit the schedule rather than assuming the original hours remain safe. See AWS guidance on supplying resources dynamically.
Free tools Windows power users keep installed
One-click scans. No signup required.
Configure permissions, timing, and recovery deliberately
Scheduler supports AWS API targets and time zones, including daylight-saving adjustments. AWS describes its delivery model as at least once, and provides configurable retry behavior. These capabilities help with scheduling, but they do not remove the need to choose suitable target actions, grant only the necessary permissions, and monitor whether the workload becomes ready as expected. AWS EventBridge FAQs.
Rank #3
- Limit IAM permissions: Give the Scheduler execution role only the actions it needs on the specific resources being managed.
- Choose the time zone intentionally: Use a named time zone when the schedule should follow local clock time; account for daylight-saving changes where applicable.
- Set retries to suit the workload: Consider what should happen if an API invocation fails and how quickly a missed start or stop must be recovered.
- Monitor outcomes: Check that the requested EC2 or ECS state change occurred and that the service is ready before use.
- Keep a manual path: Ensure operators can start the environment outside its scheduled window when a test or incident requires it.
The case study’s “zero code changes” means no application-code edit and no Lambda function for this arrangement. It does not mean zero setup, zero operational oversight, or that every application can safely be stopped and restarted on a timer.
Understand what can still cost money
Stopping an EC2 instance or setting an ECS service’s desired count to zero does not prove that all related AWS charges have ended. Storage, networking, and other allocated services may continue to incur costs. The $19.18 and $2.40 figures in the case study refer to its estimated staging compute costs, not a verified total for every line item on an account bill.
Check Cost Explorer or your billing data before and after the change, and review the resources that remain allocated against current AWS pricing. EventBridge Scheduler itself also has pricing terms: AWS’s pricing page, accessed Oct. 7, 2026, lists 14 million invocations per month in the Free Tier and uses $1 per million invocations above that amount in its pricing example. AWS says the Free Tier amount is calculated monthly across Regions except GovCloud; unused usage does not carry forward. Pricing can change, so verify the current terms. Check EventBridge pricing.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteWhen this approach is—and is not—the right fit
A few direct Scheduler API calls can be appropriate when you have a small number of resources and a predictable on/off window. Centralized or more complex environments may need a different scheduling design, more monitoring, or additional orchestration. AWS’s Instance Scheduler on AWS is a distinct solution described as supporting tagged resources and cross-account scheduling, with an architecture that uses EventBridge rules, Lambda, and DynamoDB. AWS also notes that Auto Scaling groups have their own scaling mechanisms. Those options are not the same as the four direct Scheduler schedules in the case study. See the AWS Instance Scheduler solution.
Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
