Unable to get all Windows instances on Grafana


  • I want to monitor window servers present on IBM cloud on Grafana.

  • For that I have imported a dashboard with ID 14694 and it is imported successfully but I am not able to see all the servers in the dashboard. Rather I am only able to see two server servers.

  • I clearly do not understand that why only 2 servers are listed in Grafana while others are not.

  • Given that all the servers have the same configuration, all the servers have Windows exporter installed and they are exporting the metrics through the port 9182.

  • 9182 port is listening on the server and is open and I am able to see the metrics on Prometheus.

  • I also tried to import another dashboard with dashboard ID 14510 but it is also listing the only two servers which were listing in the previous case.

  • I have installed Windows exporter on all the Windows servers which I need to monitor and they are also sending metrics to Prometheus. But I am only able to see the same two servers in the dashboard.


So that dashboard uses the server variable to fill in the drop down

label_values(windows_cpu_time_total, instance)

what is different about those 2 servers than all others based on this above query?

Hello, thank you for replying.

I have not used any query per say.

I have setup the Prometheus configuration and listed the servers like this in prometheus.yml file,

  • job_name: ‘My Windows Job’
    - targets: [‘xx.xxx.x.xx:9182’]

The Windows server jobs are up on the Prometheus and giving the metrics as well.

The issue is with the Grafana which is not reflecting all the servers but only two. (These two servers are setup the same way as well.)

Attaching the screenshot of the Grafana dashboard for the reference as well.

you have not used any query but the person that built this dashboard has. it is best to fully understand what someone else has created.

Click on the gear to the right of the save button. that should open this list and then click on variables.

there you will see servers with this query

label_values(windows_cpu_time_total, instance)

that is what populates the list of your servers. you have to understand why only 2 servers populate.

Are you seeing the other servers within prometheus?

Yes, I am seeing the following query on the dashboard.

And yes I am able to see other servers on Prometheus.

So why are they not showing up in this query

label_values(windows_cpu_time_total, instance)

can you run this query outside of grafana, in prometheus to see what it returns?