Energy optimization of OpenStack-based cloud data center infrastructures


Funding: National R&D&i Programme for Societal Challenges, RETOS-COLABORACION (MINECO)

Partners:  StackOps (MásMóvil Group), Aquads, LSI (UPM)

Duration: 2014-2016

Principal Investigator: Diego Parrilla (StackOps)

The GREENSTACK project focuses on Energy Optimization of OpenStack, providing it with awareness of the behavior of the data center to accurately anticipate actual needs.

The specific objectives of this project are:

  1. Understanding the impact of various decisions of the IaaS infrastructure in the energy consumption and the thermal behavior of servers and storage, as a function of workload and resource status and characteristics. Develop efficient models to anticipate the effect of a decision on OpenStack.
  2. Develop optimization algorithms for scheduling, resource allocation and orchestration within OpenStack, coordinated with the actuation on the cooling systems, and using the previous models to reduce the overall energy consumption of the data center.
  3. Provide a case study of business model based on close cooperation between the infrastructure and IT systems, that can be replicated for the various components, hardware and software, of the data center to reach unprecedented savings rates.