Autoscaler

The purpose of this project is to study autoscalers and more specifically the predictive autoscalers that are used by AWS and other cloud providers to scale automatically resources based on various application-dependent metrics.

The project involves two steps:

1. Study the paper A simple and effective predictive resource scaling heuristic for large-scale cloud applications

2. Implement a simulation of the heuristic - you can use SimPy simulation framework and OR Tools, MIP Solver for this purpose. It is perfectly fine if you want to simplify the forecasting model.