I would like to monitor all processes of a server, so that e.g. if there is high I/O load / network bandwidth measured by node exporter, it is possible to correlate this data with the processes "active" at that point in time.
Collecting all processes is possible already, but creates a lot of (unneeded) data. Is there a way to collect only metrics for processes which are non-idle (>= some configurable threshold)?
Hi,
I would like to monitor all processes of a server, so that e.g. if there is high I/O load / network bandwidth measured by node exporter, it is possible to correlate this data with the processes "active" at that point in time.
Collecting all processes is possible already, but creates a lot of (unneeded) data. Is there a way to collect only metrics for processes which are non-idle (>= some configurable threshold)?
Thanks!