Philip Withnall created an issue #372:
In order to be able to compare the carbon emissions from GUADEC 2020 against previous (and future) GNOME conferences, it would be really useful if some data could be collected from the servers during the conference:
- Number of machines in use (physical or VMs)
- Carbon intensity of the power supply of each machine or, if unknown, location of the machine (country or state)
- Average load at half-hourly intervals throughout the conference
- Total network TX and RX bandwidth used throughout the conference by each machine
- Total network TX and RX bandwidth used after the conference by each machine still running (e.g. for allowing people to download recorded talks)
- Whether any new hardware (servers, microphones, AV hardware, whatever) was purchased for the conference
- Change in bandwidth consumption for general GNOME services (IRC, GitLab, etc.) over the course of the conference (this one’s less important, and possibly harder since it involves changing the configuration of existing machines)
Would that be possible? The data doesn’t need to be in any particular format (as long as it’s unambiguous) — I can post-process it. I don’t need the data to be available in realtime; having a dump of it after the conference is over would be fine.
If it would help, and if you don’t have a system for it already, I can put together a script which could be run on a systemd timer to collect the time series data.
/cc @averi @barthalion