Schedule EC2 Start/Stop

Quick Answer

Reducing costs by stopping non-production instances during off-hours is fundamentally a scheduling problem, and AWS Instance Scheduler is the AWS solution purpose-built to automate exactly that: starting and stopping EC2 instances on a defined, recurring schedule so instances aren't left running, and billing, during nights, weekends, or other periods when they're not needed. What distinguishes it from the other services in this scenario is that scheduling recurring start and stop actions is its core, dedicated function, rather than a capability it happens to offer as a side effect of solving a different problem. Systems Manager Maintenance Windows can technically be used to trigger start and stop actions through automation documents, but that's a more general-purpose automation and patching tool being repurposed for scheduling, rather than a solution designed around this specific use case. CloudWatch Alarms can trigger actions based on metric thresholds, but they don't have native, built-in support for recurring time-based schedules, which is what's actually needed here. EC2 Auto Scaling, meanwhile, adjusts instance counts based on demand signals, not on a fixed time-of-day schedule, so it doesn't fit a requirement framed purely around off-hours cost savings. Whenever a question describes wanting to start and stop instances automatically on a recurring time-based schedule specifically to cut costs on non-production resources, that's the signal pointing to AWS Instance Scheduler.