This repository contains datasets, simulation code, and analysis notebooks used in the paper "Let's Wait Awhile: How Temporal Workload Shifting Can Reduce Carbon Emissions in the Cloud":

• data/*: Energy production and carbon intensity datasets for the regions Germany, Great Britain, France (all via the ENTSO-E Transparency Platform) and California (via California ISO) for the entire year 2020 +-10 days.
• compute_carbon_intensity.py: The script used to convert energy production to carbon intensity data using energy source carbon intensity values provided by an IPCC study.
• simulate.py: A simulator to experimentally evaluate temporal workload shifting approaches in data centers with the goal to consume low-carbon energy.
• analysis.ipynb: Notebook used to analyze the carbon intensity data.
• evaluation.ipynb: Notebook used to analyze the simulation results.

For executing the code you need to install the libraries listed in environment.yml, e.g. by using a conda environment.

If you use any datasets or code from this repository, please reference our publication:

• Philipp Wiesner, Ilja Behnke, Dominik Scheinert, Kordian Gontarska, and Lauritz Thamsen. "Let's Wait Awhile: How Temporal Workload Shifting Can Reduce Carbon Emissions in the Cloud" In the Proceedings of the 22nd International Middleware Conference, ACM, 2021.

BibTeX:

@inproceedings{Wiesner_LetsWaitAwhile_2021,
  author={Wiesner, Philipp and Behnke, Ilja and Scheinert, Dominik and Gontarska, Kordian and Thamsen, Lauritz},
  booktitle={Middleware'21: 22nd International Middleware Conference}, 
  title={Let's Wait Awhile: How Temporal Workload Shifting Can Reduce Carbon Emissions in the Cloud}, 
  publisher = {{ACM}},
  year={2021},
  doi={10.1145/3464298.3493399}
}