In the context of global warming, reducing the carbon emissions of computer infrastructures is a top priority. Datacenters are responsible for a large fraction of the carbon emissions of IT.
To reduce the carbon footprint of datacenters, actions at multiple levels need to be taken. The way of managing and using the hardware should change [1]. However, efforts made only at the hardware level will not be enough to solve the challenges we are facing: software design needs to be made carbon aware [2]. A good example that illustrates the challenges we are facing is reliability and storage. To ensure the availability of data in the Cloud despite crashes, data are replicated on different storage devices. Encoding techniques can also be used as a means to reduce the storage capacity used for fault tolerance [3]. These techniques have different advantages and drawbacks but have not yet been analyzed from the point of view carbon footprint. From this point of view it is difficult to identify a priori a solution that is better than the others: Some solutions use more storage space and less computing resources, while it is the opposite for others, and it is difficult to know what is best.
Considering the reliable storage of data in datacenters, the goal of this internship is to analyze the carbon footprint of different approaches and, if possible, to propose a new approach that would reduce the carbon emissions.
The intern will be in charge of the following tasks:
Studying recent publications and collecting information about the carbon footprint of current storage systems.
Running experimental campaigns on our large-scale private Cloud platform, to analyze the energy/performance trade-off provided by different software solutions for the reliable storage of data.
Analyzing the results to propose a strategy to reduce the carbon footprint of reliable storage approaches
[1] Tomlinson, Amanda, and George Porter. “Something Old, Something New: Extending the Life of CPUs in Datacenters.”
[2] Anderson, Thomas, et al. “Treehouse: A Case For Carbon-Aware Datacenter Software.” arXiv preprint arXiv:2201.02120 (2022).
[3] Kim, Athlur, et al. “Morph: Efficient File-Lifetime Redundancy Management for Cluster File Systems” (2024)
The intern will join the Informatics Laboratory of Univ. Grenoble Alpes (LIG), one of the largest laboratory in Computer Science in France. Univ. Grenoble Alpes is one of the 50 best universities in the world for computer science according to international rankings.
More specifically, the intern will work in the KrakOS research team, that focus on the efficiency and robustness of systems. KrakOS is one of the top system research teams in France and in Europe, with publications in the most selective conferences of the domain.