Tech Showcase: Resource Central: Toward Intelligent Clouds

This project is exploring ways to create intelligent and efficient cloud platforms by leveraging machine learning (ML) techniques. Our first foray in this direction has produced Resource Central (RC), an ML and prediction-serving system that provides intelligence to the various resource managers in a cloud platform. Specifically, RC collects telemetry from virtual machines and servers, learns from their prior behaviors and, when requested, produces predictions of their future behaviors. RC is currently in production within the Azure Compute Fabric. Our ongoing research effort is exploring other architectures and approaches to integrating ML into cloud systems. In particular, we are currently investigating how to broaden the set of management tasks that RC can support. Our poster will highlight our vision for intelligent and efficient cloud platforms, describe potential approaches for integrating ML, and provide an overview of RC and its initial results.

Date:
Haut-parleurs:
Anand Bonde

Taille: Microsoft Research Faculty Summit