I've built a dashboard on grafana using sever metrics, including ones exposed by the windows_exporter. I've configured it to send cpu,memory and process information for now.

I included a query that's supposed to display the cpu usage for a single process. It worked at first but for some reason it stopped working. I don't believe I changed anything.

The query is as follows:

    100 * (
    sum(rate(windows_process_cpu_time_total{process="OpcUaLauncher"}[1m]))
) / (
    sum(rate(windows_cpu_time_total[1m]))
)

When checking the parts seperately, it turns out the issue is the top part (process part).

Furthermore, I cannot get any process information to be output in grafana at all, while the "regular" cpu and memory queries still work. This indicates to me that the data source confriguration is fine.

On my prometheus localhost:9090 tab, the query runs just fine, so the query should be correct. Again, it worked in grafana as well at some point. When looking at the metrics tab the metric is also cleary there with an accompanying value.

My only conclusion is that something is going wrong with grafana fetching metrics from the prometheus server, but the process metrics specifically. If anybody has some more insight into what the problem could be that would be great!