deployer auto scaling and destruction

[enterprise-1701-d]

Is your feature request related to a problem? Please describe.
There are either too much or too less deployers available. During breaks, nights, weekends they are either running of have to be stopped and started - that is quite expensive from resources or manual activities perspective.

Describe the solution you'd like
The number of deployers are a static number today. It would be nice to have deployers created and configured based on the need and also destroyed if not needed. So the idea is to have a very small watcher VM just checking that there are requests from the pipelines which are queuing. With pending requests a deployer is created fresh with the sufficient size and destroyed after 1hour of idle time. So you have an automatic scaling.

Describe alternatives you've considered
Yes, create a set of deployers and send them to sleep and start them if your pipeline is queuing.

[jhajduk-microsoft]

@PankajMeshramCSA @KimForss - The deployer should be created per region and contains the configuration files for the environments. Policy, auto start and stop, etc can be used to start and deallocate VMs on a schedule. Deallocated VMs incur no cost and there will be no manual intervention. Having a watcher VM introduces additional cost. I believe this is over engineering a simple issue. LMK what you think.