codecarbon VS scaphandre

The underlying calculations apparently depend on the following CSV: https://github.com/mlco2/codecarbon/blob/master/codecarbon/d...

Two problems I see here

1. They are using AWS, which is not in the data set

2. The data set seems questionable to me. The GCP Iowa data center has some of the best renewable usage, are they using the data center total (a larger data center with more renewable can still have more total emmissions vs only 8% of the compute I use should be included in the calculation. Either way, the underlying data for the calculation is out of date

I really like codecarbon to monitor my power usage: https://github.com/mlco2/codecarbon

Scaphandre!

Hello, Powertop is the bare minimum. I use it many time. But now all my Power consumption rely on Scaphandre. It’s a huge tool! I use it for 2 years now, and double check with a Wattmeter to see any gap (there is no gap). https://github.com/hubblo-org/scaphandre

Perhaps distributed systems traces could be extended to attach energy usage alongside other metadata (like request duration).

See also this HN post[1] re: an environmental impact reporting dashboard built by the BBC's R&D department. In that example, a system-level resource monitoring agent called 'scaphandre'[2] was used -- I think both are valid approaches (a tracing-based approach would allow discovery and optimization of energy-inefficient requests; a system-level approach provides good fleet-wide energy usage visibility).

[1] - https://news.ycombinator.com/item?id=30320147

[2] - https://github.com/hubblo-org/scaphandre/