How I monitor my Windows machines using Grafana
Last updated
Was this helpful?
Last updated
Was this helpful?
Well it's always a good idea to have something to trace back if either my machines is acting weird or broken down for no reason.
Basic concept of Windows Services.
Know how to use Windows Terminal - Powershell or CMD.
Have an editor installed, either Notepad++ or Visual Studio Code (use Notepad? Fine, but beware of the file encoding and extension).
I'll cover this in another post, but yeah, I'm gonna use the Cloud version of Grafana since I don't trust any of my ISPs or Electrical Providers for my homeserver. The Free-tier of Grafana Cloud seems to be more than enough for me at the time.
It's great, almost perfect replacement for Prometheus as we can now ship our logs and metrics directly to our cloud hosted Prometheus data source without the need of prometheus.
The default installation of Grafana Agent already included Windows Events logs, but if you need to collect other logs as well, simply add the follow configuration to the logs section of your grafana-agent.yml
:
Save the file and restart the Grafana Agent service via service.msc
.
Head over to their repo above and follow their installtion instruction (using their sample configuration, we'll edit them after). After having the exporter installed, let's export some metrics shall we?
Add the following to the OhmGraphite.exe.config
- located in the same location of the exe file.
Save it and finally restart the OhmGraphite service via services.msc
.
Go to your Grafana Cloud instance's dashboard - Click on New then Import, paste the ID and voila.
Those forks at OhmGraphite - did an amazing jobs of collecting not only CPU/MEM/... things that Grafana Agent can cover but also Power and GPU, and I really need these metrics.
Head over to and copy the ID.