This project aims to provide a clear defined API that offers ways of interacting with a FaaS platform (i.e., deploy function) and expose runtime metrics (i.e., recent invocations) to develop strategies on managing function deployments (i.e., scaling).
EdgeRun aims to offer two different (ready-to-use) implementations that stem from completely different backgrounds:
- galileo-faas: implements this project for a real-world testbed based on Kubernetes and galileo-experiments.
- faas-sim: offers a trace-driven, event-based simulation for FaaS.
The main use case of this project is to clearly define our vision of (1) interacting with a FaaS platform and (2) exposing runtime metrics to develop novel system components that revolve around scaling, scheduling and load balancing.
This translates to the following concrete use cases:
- You want to deploy, shutdown and invoke a running FaaS platform (with
FaaSSystem) - You want to implement a scaling/scheduling or load balancing solution that uses runtime metrics (with
PlatformContext)