feat: Add Helm chart and Kubernetes tutorial #72
Conversation
| Create a configmap to store the Grafana dashboard. | ||
|
|
||
| kubectl create configmap scaphandre-dashboard \ | ||
| --from-file=scaphandre-dashboard.json=docs_src/tutorials/grafana-kubernetes-dashboard.json |
There was a problem hiding this comment.
This was the best way I could find to install the dashboard. Happy to move the files if they would be better somewhere else.
There was a problem hiding this comment.
Maybe we could find a better location later. I guess this is fine for now.
There was a problem hiding this comment.
Yes I actually wanted to try loading the dashboard from Grafana Cloud here
https://grafana.com/grafana/dashboards/13845
But I ran into a bug with the Grafana chart.
grafana/helm-charts#200
|
Thanks a lot for the PR ! |
| First we install Scaphandre which runs as a daemon set which creates a pod on | ||
| each node for collecting the metrics. | ||
|
|
||
| helm install scaphandre helm/scaphandre |
There was a problem hiding this comment.
Does that suppose we rely on the local chart ? Maybe we should add the step to put the chart on the right location first ? (may seem obvious but this is to avoid confusing helm users used to download remote charts)
There was a problem hiding this comment.
Yes its using the local chart but that's not ideal.
The central Helm charts repo is now deprecated and not accepting new charts. https://github.com/helm/charts#helm-charts
The new pattern is to have a charts repo per org e.g. https://github.com/grafana/helm-charts and then link it to https://artifacthub.io/ for discoverability.
It should be possible to use something like GitHub Actions to push new versions of the chart to the repo. Happy to help with that if I can.
There was a problem hiding this comment.
This is fine as a first version.
We may create an FR after that to setup a helm chart repository for scaphandre.
|
I've tested it on a local cluster and the whole process works like a charm ! Thanks a lot. 😃 I'll integrate the recent fix I made for the CI job in a local branch and merge from there. EDIT: I've just noticed that some targets of the prometheus deployment and down with "Get "http://192.168.1.31:9100/metrics": context deadline exceeded":
It is completely unrelated to scaphandre but if you have some information about that, I'd be interested to fix it in the process. |
Related to #29
This adds the Helm chart as well as a tutorial for installing it as well as Prometheus and Grafana.
The dashboard file is pretty huge but here are the results from my test cluster.