Hybrid Performance-aware
Power-capping Orchestrator

Get Started Our Projects

Energy proportionality
for on-premise applications

HyPPO is an orchestrator for microservices designed to guarantee the workloads performance while minimizing unnecessary power usage

  • Supports GNU/Linux, Docker containers and Kubernetes clusters.
  • Targets Cloud and On-Line Data Intensive workloads.
  • Achieves autonomic run-time management of performance and power.

Features

HyPPO is based on the concept of distributed ODA loop. To reach its goals HyPPO Observes in real-time the behaviour of the kubernetes cluster. The resulting data is leveraged by a controller which Decides how to allocate power respecting the performance needs of each container. Finally, an Act phase enforces power and performance on the cluster.

System monitoring

Measure power consumption, resource usage and performance metrics for each container.

Cluster view

Filter, segment and show real-time metrics from the single container to the entire kubernetes cluster.

Autonomic control

Guarantee latency and CPU request usage saving power in an energy proportional way.

Power management

Leverage hardware and software tools available in current servers to handle power allocations.

Technology

HyPPO is built on top of industry standard technologies and on top of established open source projects to deliver performance, precision, power-efficiency and an enhanced end-user experience

Running projects

HyPPO is continuously evolving! Here you can find some of the projects we are working on

performance and power monitoring

Distributed ODA loop

Latency awareness

Predictive controllers

Multi-target actuators

FPGA as a microservice

Publications

Asnaghi, Amedeo, Matteo Ferroni, and Marco D. Santambrogio. "DockerCap: A Software-Level Power Capping Orchestrator for Docker Containers." Computational Science and Engineering (CSE) and IEEE Intl Conference on Embedded and Ubiquitous Computing (EUC) and 15th Intl Symposium on Distributed Computing and Applications for Business Engineering (DCABES), 2016 IEEE Intl Conference on. IEEE, 2016.

Arnaboldi, Marco, Matteo Ferroni, and Marco D. Santambrogio. "Towards a performance-aware power capping orchestrator for the Xen hypervisor." ACM SIGBED Review 15.1 (2018): 8-14.

Rolando Brondolin, Tommaso Sardelli, and Marco D Santambrogio. Deep-mon: Dynamic and energy efficient power monitoring for container-based infrastructures. In Parallel and Distributed Processing Symposium Workshops, 2018 IEEE International, pages 676–684. IEEE, 2018.

Marco Arnaboldi, Rolando Brondolin, and Marco D Santambrogio. Hyppo: Hybrid performance-aware power-capping orchestrator. In Autonomic Computing (ICAC), 2018 IEEE International Conference on. IEEE, 2018.

Rolando Brondolin, Marco Arnaboldi, Tommaso Sardelli, Sara Notargiacomo, and Marco D Santambrogio. Energy efficiency for autonomic scalable systems: research objectives and preliminary results. In Research and Technologies for Society and Industry (RTSI), 2018 IEEE 4th International Forum on. IEEE, 2018.

Our Team

Rolando
Brondolin

PhD Student @ Polimi

Marco
Arnaboldi

Master of Science @ Polimi

Tommaso
Sardelli

Master Student @ Polimi

Marco
Bacis

Master Student @ Polimi

Samuele
Barbieri

Bachelor Student @ Polimi

Andrea
Strada

Master Student @ Polimi

Daniele
Rossetti

Master Student @ Polimi

Sara
Notargiacomo

Tech Transfer Manager @ NECST

Marco D.
Santambrogio

Associate Professor @ Polimi

Contact Us

If you want to improve the energy efficiency of your applications, if you want to help us to shape the future of energy proportional datacenters or if you are just interested in our research, drop us a line!

Address

Via Ponzio 34/5, 20133 Milano MI, IT

Proudly developed
by the NECSTLab guys

Email

hyppo at necst dot it