Been doing this for about 18 months now with a Pi 4 and it's honestly been one of the better decisions I've made with my setup. Running VenusOS on it to pull data from my Victron kit, then pushing everything into InfluxDB with Grafana dashboards on top.
The long-term data has been genuinely useful — spotted that my Fogstar batteries were consistently underperforming in cold snaps before I'd have otherwise noticed. Would have missed that completely without the historical graphs.
A few questions for anyone else doing something similar though:
- How are you handling SD card wear? I've read about cards dying after a year or two of constant writes. Currently running off a USB SSD instead but wondering if there's a better approach
- Are you logging to the cloud as well, or keeping it all local? I'm a bit wary of sending my energy data off-site but local-only means I can't check in remotely
- What retention period are you keeping data at full resolution? I'm keeping 1-minute intervals for 90 days then downsampling, but not sure if that's overkill
Also curious whether anyone's gone further and used the historical data to actually tune their system — like adjusting charge profiles seasonally or predicting when you'll run low based on weather forecasts. Feels like there's a lot of potential there that I'm barely scratching the surface of.
Running a tiny house setup with EV charging thrown in, so the load patterns are pretty chaotic and I feel like there's proper insight buried in the data somewhere if I knew how to dig for it.