View Multiple Java process metrics on Windows in netdata

Hello

I have netdata running and monitoring some windows servers, using the Prometheus windows exporter and the go.plugin on netdata. I am trying to configure netdata to display CPU / memory usage for 8 individual Java processes on my windows servers. But netdata lumps all 8 java processes into one metric. How do i get netdata to split up the metrics into 1 per process?

On the Prometheus windows exporter config i have this regexp to pickup java processes:

–collector.process.whitelist=“java.*”

You can see here the 8 separate Java metrics (see on http://:9182/metrics)

windows_process_cpu_time_total{creating_process_id=“3064”,mode=“privileged”,process=“java”,process_id=“8788”} 5.4375
windows_process_cpu_time_total{creating_process_id=“3064”,mode=“user”,process=“java”,process_id=“8788”} 30.75
windows_process_cpu_time_total{creating_process_id=“384”,mode=“privileged”,process=“java”,process_id=“4252”} 20.859375
windows_process_cpu_time_total{creating_process_id=“384”,mode=“user”,process=“java”,process_id=“4252”} 240.296875
windows_process_cpu_time_total{creating_process_id=“4312”,mode=“privileged”,process=“java”,process_id=“10188”} 51.984375
windows_process_cpu_time_total{creating_process_id=“4312”,mode=“user”,process=“java”,process_id=“10188”} 265.25
windows_process_cpu_time_total{creating_process_id=“4548”,mode=“privileged”,process=“java”,process_id=“2636”} 19.875
windows_process_cpu_time_total{creating_process_id=“4548”,mode=“user”,process=“java”,process_id=“2636”} 240.0625
windows_process_cpu_time_total{creating_process_id=“5292”,mode=“privileged”,process=“java”,process_id=“5236”} 6.9375
windows_process_cpu_time_total{creating_process_id=“5292”,mode=“user”,process=“java”,process_id=“5236”} 61.421875
windows_process_cpu_time_total{creating_process_id=“6028”,mode=“privileged”,process=“java”,process_id=“10528”} 6.65625
windows_process_cpu_time_total{creating_process_id=“6028”,mode=“user”,process=“java”,process_id=“10528”} 63.03125
windows_process_cpu_time_total{creating_process_id=“6512”,mode=“privileged”,process=“java”,process_id=“1644”} 294.578125
windows_process_cpu_time_total{creating_process_id=“6512”,mode=“user”,process=“java”,process_id=“1644”} 174.734375
windows_process_cpu_time_total{creating_process_id=“6672”,mode=“privileged”,process=“rvdaemon”,process_id=“6000”} 5.515625
windows_process_cpu_time_total{creating_process_id=“6672”,mode=“user”,process=“rvdaemon”,process_id=“6000”} 7.28125

but in netada dashboard i see this in attched pic

Hi @mikewilson ,

Welcome to the community and thanks for raising this question to us.

At the moment what your asking isn’t possible using the local Agent dashboard - from your screenshot I’m assuming you haven’t tried this on Netdata Cloud. We will be working on revamping the Agent dashboard but we don’t have yet an ETA for when this will be available.

If you take a look at Netdata Cloud, you can check our Windows room in the Netdata demo space here, you will see that the metric windows.processes_cpu_utilization has the processed identified as a dimension and you can group them both on this and on node or system(instance) so we get a full breakdown of the processes by windows server being monitored - please note that for node you need to have these as Virtual Nodes, config details on our docs here

Example of the group by selections

Result

If you can, please give it a try on Netdata Cloud and see if you can achieve your desired result.

Regards,
Hugo