Enforcing on/off policies – scheduled and on-demand
Since you're reading a cloud cost-optimization book, I think you may be one of the turn off the tap when you don't need water club members.
The same best practice could and should be used with cloud resources: if you don't use a service, database, or VM, then that resource should not be running without purpose.
Usually, we can identify two types of on-off policy, as follows:
- Scheduled: We use this policy when you set a power-on time and a power-off plan, and it's always the same.
- On-demand: We use this policy when the resources are always powered off and the user starts them up only when needed.
Of course, the on-demand approach is the best, when speaking in terms of cost control, because you use the cloud resources only when needed, and for the right time. The scheduled approach, instead, may leave the resources on even if unused.
Non-production environments (test, development...
