How usage-based resource management works ?

Hello, how Railway manages resource scaling ? like how are vCPU and RAM scaled up and down automatically without downtime ? How does Railway estimate the amount our app needs ?
Thanks.

(I don't have a project yet...)

there's no estimations for resource allocations and there's no scaling, the docs use misleading words, your service gets access to 8 vcpu and 8gb of ram and is free to use as much or as little as it needs, then you are bill accordingly for what you did use

ok so when our services doesn't use it, this free resource is used by other users apps ?

each service gets dedicated resources

I don't see how it can be profitable if a lot of users are using the lowest range

🤷