19-22 May 2014
Helsinki University, Main Building, Helsinki, Finland
Europe/Madrid timezone

Efficient Management of an OpenStack Cloud Infrastructure Through Multiobjective Programming

Not scheduled
Helsinki University, Main Building, Helsinki, Finland

Helsinki University, Main Building, Helsinki, Finland

Poster

Speaker

Miguel Angel Diaz (CETA-CIEMAT)

Description of content and intended audience</br>- the outcome you expect to achieve.<br/>

Description of contents: Our experiments analyse the influence of the ACPI module, the boot costs of a node configured with the different ACPI sleep status, virtual machine migration costs, comparisons between simulated and real cloud infrastructures and comparisons among the different alternatives explained in this poster.

Audience: System administrators, developers and data center directors

Outcome: The aim of this poster is to present our work, show our interest for energy and cloud technology, and meet new contacts.

Printable summary: this is the only </br> section of the abstract that will <br/>be published in the Book of Abstracts.

Cloud Computing as IAAS (Infrastructure as a Service) has gained a rapid
acceptance in the IT world producing implantations of clouds
inside data centers. Focused on large cloud infrastructures, hundreds of machines
are working together to offer users a base to execute their virtual machines.
However, these cloud infrastructures consume large amounts of electrical
energy, contributing to high carbon footprints to the environment. In addition,
users expect the best reliability and availability of the cloud services when their
business are depending on those services. Therefore, we need a Green Cloud
Computing solution which does not only minimize energy consumption, but also
achieves the efficient processing and utilization of cloud computing infrastructures.

In this poster, we work with a real cloud architecture based on the OpenStack
platform to achieve a solution through real infrastructures
instead of simulators. Our experiments analyse the influence of the Advanced Configuration
and Power Interface (hereinafter ACPI) module, the boot costs of a
node configured with the different ACPI sleep status, virtual machine migration
costs and comparisons among the different alternatives explained in this poster (without
threshold, with one threshold and with two thresholds). The result is the development
of a set of daemons which minimizes the energy consumption, and also
achieves the efficient processing and utilization of a cloud computing infrastructure.

Primary author

Miguel Angel Diaz (CETA-CIEMAT)

Co-authors

Dr Abel Paz (CETA-CIEMAT) Dr Miguel Angel Vega (University of Extremadura)

Presentation Materials

There are no materials yet.