Today at the #wendelstein #stellarator #fusion #experiment:
Documentation! And learning to not just write it, but also test it.
TL;DR, let someone else go through it. The second pair of eyes makes very different mistakes than yours.
We are setting up a permanent server for the sensor boards that we have been working on for the past few months - finally, a proper place to put the #grafana and #postgresql #timescaledb instances and the #ruby backend code.
In order to make this repeatable, we wrote a page or two in the #gitlab wiki of the project about how to go about it, and our supervisor set up the server by following said wiki...
And immediately things didn't line up quite right.
The server didn't have internet, so we had to figure out how to download the .deb packages instead - same for the Ruby bundler packages!
Then some issues with the versioning, and setting up Timescale...
Nothing *insurmountable*, but if we hadn't been there in person, it might not have worked as smoothly.
So, even if you write documentation, make sure someone else tries to go through it on their own, and figure out where things get stuck!
Pictures of the new Grafana #dataviz panels coming up soon - I gotta say, Grafana 9 looks very lovely!
Work is proceeding nicely at the #wendelstein #fusion #experiment!
The arc detection boards are due to being tested tomorrow, after the #data #server was successfully moved to a permanent virtual machine! More news on that soon.
For now, have another screenshot of a new, gorgeous #grafana dashboard for the devices. We figured out that you can use repeats on entire rows, which is wonderful for what we want to do!
@xaseiresh great to see this as another case of open source software coming in handy for fusion :) I'll have to look into grafana myself, do you have any resources besides what's available on their website to recommend? :)
@StanczakDominik
Grafana itself is quite easy to work with, I think the [official documentation](https://grafana.com/docs/grafana/v9.0/getting-started/build-first-dashboard/) is nicely fleshed out already :)
The more important part is the data source that you're using!
This project uses #timescaledb because it integrates very nicely with Grafana and has a few other useful abilities.
[We wrote a blog about how we use it a while ago!](https://www.timescale.com/blog/using-iot-sensors-timescaledb-and-grafana-to-control-the-temperature-of-the-nuclear-fusion-experiment-in-the-max-planck-institute/)
For server monitoring you might instead want to look into InfluxDB or similar. Grafana mostly just plots what you give it :)
