In this article I want to propose a simple way to do a Software Life Cycle Assessment for a piece of code. While explaining important concepts and tradeoffs that seem sensible for a little project. Of course things could be done differently and I am happy to discuss them in more detail as this is just a first draft of how things could be done. Please head to https://github.com/green-coding-solutions/SLCA-Code for all source code and discussions. Please also feel free to contribute to the code through pull requests. As Green Coding Solutions we have been working on this topic for some time and a more academic discussion can be found in our chapter in the Transparency for Software Climate Impact publication (Part of the Shaping Digital Transformation for a Sustainable Society publication).

Getting power usage data in cloud is a tough topic and one of the reasons projects such as Cloud Carbon Footprint exist. Typically cloud vendors, especially hyperscalers, do neither supply energy consumption data of the whole machine from PDUs or similar, nor do they provide access to CPU internal energy data like for instance RAPL.

We are happy to announce a major update to our Eco-CI tool, a plugin designed to estimate and reduce the energy consumption of continuous integration pipelines. Previously available exclusively for GitHub Actions, we are excited to announce the integration of Eco-CI with GitLab pipelines. We hope this will encourage Gitlab developers to measure their carbon emissions and make informed decisions regarding sustainable software processes.

At Green Coding Solutions (GCS), one goal is to enable reproducible runs on our cluster. An important step towards accurate measurements was the creation of NOP Linux, our custom Linux distro that disables as many background processes as possible to avoid interruptions during measurements. Another crucial step was ensuring the reliable operation of the PowerSpy2, so we could measure the entire power consumption.

The long awaited feature to make comparisons with the Green Metrics Tool is finally live 🔥

Our measurement workflow now includes phases (Baseline, Idle, Installation, Boot, Runtime, Remove) which makes our tool compatible with the requirements specified by the Blue Angel for Software.

Green Coding on Mac

When we started with the Green Metrics Tool (GMT) we set out to develop tooling to enable developers, decision makers and the wider public to measure and see the impact software is having on the environment.

As we mentioned in a previous blog article, we’ve been thinking of ways to make CI pipelines more energy efficient. One idea we had was to give developers a tool which allows them to measure the energy used by their CI runs.

Many of our developers work on macOS exclusively. In order to create a nicer development experience, we are currently undergoing some work to port our Green Metrics Tool to macOS.

One question we’ve been tinkering around with here at Green Coding Solutions is how can we make CI pipelines around the world a little bit greener.

In the green software community we see very often that people use the TDP of the processor as a metric to estimate how much energy the CPU will consume for a specific workload.