Can't get TrueNAS dashboard working with graphite exporter

Hey,

I’m trying to get Prometheus, Grafana, and graphite-exporter working together to monitor my TrueNAS (hostname: kjnas), and I need some help. I was able to setup monitoring with node-exporter on my separate Linux hosts, so Grafana and Prometheus are talking.

My current setup is an LXC container on Proxmox that is running Debian 12 and has Docker and Docker Compose installed. The IP of that machine is 192.168.50.48.

I put all of my configs and metrics snippet in a single Pastebin link as I can only post 2 links:

And all images in one Imgur album:

Versions

  • Grafana (Grafana v10.4.3 (0bfd547800))
  • Prometheus (2.52.0)

Truenas Exporter setup: Image “1” in the imgur album

I can see my “truenas” Prometheus job as a target, and it’s getting metrics. A snippet of the metrics I can be seen in my pastebin link.

I’m trying to use this dashboard ID: 19661

When I import that dashboard and select Prometheus as a data source, I don’t see any data. On the top, there is a “hostname” dropdown, which is not displaying anything:
Image “2” in the imgur album

So I assumed it had to do with how that variable is configured. I found the settings for the variable, but sadly, I’m not smart enough to figure out how to fix this: Image “3” in the imgur album

Any advice?

Did you follow all steps described in manual?

Yes, I tried. If use mappings provided in that repo, I get this error when I’m trying to start the container:

graphite | ts=2024-07-15T18:42:51.238Z caller=main.go:96 level=error msg="Error loading metric mapping config" err="yaml: line 3: did not find expected '-' indicator"

^^^^
EDIT: Forget this, it was just YAML thing. There was a incorrect indentation, I fixed it so now we are back to the original issue.

Also the screenshot from TrueNAS doesn’t look at all like mine looks.

Then I continued my journey to this post:

and I used the mappings provided from there. These resulted in the same problem as described in my original post.