I’m very new to grafana - as in about an hour ago. My org has a windows server down. I was given the login to the grafana display - from a rhel 8 server running prometheus (or does the data for all of our servers come from the node exporter to prometheus, which reports to grafana - still figuring that out). Anyway a server is down and the display goes from red to green to red back to green. I go in to edit that panel and originally there was no value under “Critical” in the “threshold” section. I put a 0 in there and the panel went red. But after a few minutes it goes right back to green. Yet the server is still down. I can’t seem to figure out what’s causing that. Can anyone point me in the right direction?
I can’t find any difference in the JSON for this down server and any of the other Windows servers configured. Could that mean that none of them will display correctly?
Hello @timothyriley2civ
After analyzing the issue, I believe I have identified the underlying problem. Please find below the recommended steps that may help resolve it.
If the issue persists, kindly provide additional details and screenshots to help us better understand and diagnose the problem.
Step 1. Check Prometheus target status
Go to http://<prometheus-server>:9090/targets and make sure your Windows server target shows as DOWN when it’s actually down.
Step 2. Update your Grafana query
up{instance="<windows-server>"} or vector(0)
This makes sure missing data counts as 0 (down).
Step 3. Set correct panel thresholds
In Grafana panel thresholds, set:
Critical = 0 (red)
OK = 1 (green)
Step 4. Configure alerting for no data
In alert rules, set No Data & Error Handling to treat missing data as Critical (alert).
Hello @timothyriley2civ
Just checking in—did this solution resolve your issue, or are you still experiencing the same problem?