# How to Automatically Shut Down Non-Production Environments at Night (2026) Canonical: https://www.pointfive.co/guides/how-to-automatically-shut-down-non-production-environments Plan non-production schedules, estimate avoidable running costs, and account for storage, commitments, restart rules, and operating costs. By: PointFive Team Published: 2026-06-02 Updated: 2026-06-02 [All guides](https://www.pointfive.co/guides) PointFive Team June 2, 2026 2 min read Scheduling dev and staging resources can reduce unused running time. Start by identifying when each environment is needed, who owns it, and which dependencies must remain available. ## Estimate the opportunity from your schedule An environment running continuously is active for 168 hours in a seven-day week. A Monday-to-Friday, 9am-to-6pm schedule uses 45 hours. That leaves 123 hours, or about 73%, outside the proposed schedule. This is an illustrative reduction in running hours, not an observed average or a promise of equivalent bill savings. Calculate the avoidable running charges separately from storage, backups, networking, licenses, and commitments. A commitment payment may continue even when a covered resource stops. Verify the change against actual billing after deployment. ## Choose an implementation Instance Scheduler on AWS provides schedules for supported EC2 and RDS resources. Its [schedule reference](https://docs.aws.amazon.com/solutions/latest/instance-scheduler-on-aws/schedule-reference.html) also supports an optional EC2 instance type per period: a running instance can be stopped and restarted to apply a different type. That setting does not resize RDS instances. The solution has [operating costs](https://docs.aws.amazon.com/solutions/latest/instance-scheduler-on-aws/cost.html) for AWS services; do not treat it as a free service or assume only Lambda charges apply. Custom automation can coordinate dependencies and team-specific exceptions. Account for implementation, permissions, monitoring, retries, and maintenance when comparing it with a packaged scheduler. PointFive OS Coworkers support customer-defined automation using connected context and authorized tools. Actions that change infrastructure require customer-authorized write access and approval before execution. The resources and operations available depend on the connected services and permissions. ## Check what stopping actually does For [EBS-backed EC2 instances](https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/Stop_Start.html) , stopping preserves attached EBS volumes but does not preserve ordinary RAM or instance-store data. Stopping is not a substitute for backups. Storage and other associated charges can continue. [RDS stop/start](https://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/USER_StopInstance.html) has engine and configuration restrictions. A stopped DB instance is automatically restarted after seven days, and storage charges continue. Coordinate recurring schedules with that behavior. [Aurora Serverless v2 automatic pause](https://docs.aws.amazon.com/AmazonRDS/latest/AuroraUserGuide/aurora-serverless-v2-auto-pause.html) requires a supported engine version and a minimum capacity setting of zero. Not every Aurora Serverless v2 configuration can pause. Check connection behavior and resume latency before relying on it. For Kubernetes, reducing replicas or nodes depends on workload controllers, minimum capacity, disruption budgets, and system dependencies. Define a tested startup sequence instead of assuming an entire cluster can always scale to zero. ## Roll out with an owner and an override - Select a non-critical environment and document dependencies, timezone, and working hours. - Confirm the required permissions and exclude production resources explicitly. - Test shutdown, startup, data persistence, health checks, and the manual override. - Notify the owner and monitor missed starts or unexpected restarts. - Compare actual charges and availability with the baseline before expanding the schedule. Sources checked September 17, 2026. ## Frequently asked questions ### What percentage will scheduling save? There is no universal percentage. Estimate avoidable running charges from the proposed schedule, then account for costs that continue while resources are stopped. ### Can Instance Scheduler change instance types? It supports scheduled EC2 instance-type changes. The same setting does not resize RDS instances. ## Related reading [All guides](https://www.pointfive.co/guides) - ### [Best AWS Cost Optimization Tools (2026): Native vs Third-Party, Compared](https://www.pointfive.co/guides/best-aws-cost-optimization-tools-2026) October 9, 2026 - ### [Best Cloud Cost Optimization Tools (2026): An Honest, Comparison-Driven Guide](https://www.pointfive.co/guides/best-cloud-cost-optimization-tools-2026) June 2, 2026 - ### [Cloud Financial Management in 2026: From Reporting to Operating](https://www.pointfive.co/guides/cloud-financial-management-2026) July 30, 2026 --- Source: the public page above. Product screenshots and illustrative interfaces are examples, not live customer data.