Reputation: 513
I want to integrate a Load Balancer for vertical Scaling. As i read, it´s possible to setup the load Balancer that point every User to the next instance.
But i thought about to do (for example) 5 VM Instances. When the first 100 Users are on Instance, the load Balancer start the next instance and so on and on ... After x Minutes if a server is unused it will turn off again until the next peak. (Maybe a Script that creates instances and delete them?)
Is that a possible and common scenario or am i completely wrong? In all the description everyone say it´s scaleable, but i don´t want to sit next to the Usage Monitor to start and stop instances.
The Keypoint is to reduce server coast in times when it´s "unused" but be able to handle peaks when i´m "out of office".
Upvotes: 0
Views: 245
Reputation: 983
There's a new feature in GCE 'Compute Engine Autoscaler', note that this is a Limited Preview and you need to apply for it.
Upvotes: 2