Comprehensive and Detailed Explanation From Exact Extract: The most effective way to achieve significant cost savings for non-production environments that are only needed during business hours is to stop the instances when they are idle. When an instance is stopped, you are only charged for the persistent disks and not for the CPU/RAM.
Instance Schedules are the native, managed Google Cloud feature designed specifically for this purpose:
automating the starting and stopping of Compute Engine instances on a recurring schedule.
* Option D (Use Instance Schedules to stop the development and acceptance environments after office hours, and start them just before office hours) directly addresses the need for cost optimization during idle times with the recommended native solution.
* Option A (Spot VMs) are cheap but can be preempted at any time, which isn ' t suitable for critical during office hours environments.
* Option B (Autoscaling) scales down when demand is low, but typically won ' t scale completely to zero instances, and will still incur minimum running costs.
* Option C (Custom Shell Script) is a manual, non-native solution that requires maintenance and is more complex than the built-in Instance Schedules.
Exact Extract Principle:
From the Google Cloud documentation on Compute Engine Cost Optimization :
" Instance schedules allow you to schedule when Compute Engine virtual machine instances start and stop.
Use instance schedules to automatically start and stop your instances at predefined times. This is a highly recommended practice for cost optimization of non-production environments (development, testing, or staging) that are not needed 24/7, as you are not billed for the instance ' s machine type when it is stopped. " Reference: Google Cloud Compute Engine Instance Schedules Documentation and Compute Engine Pricing
/Cost Optimization Guide.