I've noticed that for some metrics, Prometheus node_exporter will use one
metric name and separate labels. For example, CPU usage has a "mode" label:
node_cpu_seconds_total{cpu="0",mode="idle"} 320045.23
node_cpu_seconds_total{cpu="0",mode="steal"} 0
node_cpu_seconds_total{cpu="0",mode="system"} 3703.1
node_cpu_seconds_total{cpu="0",mode="user"} 4713.87
Whereas for other metrics, it uses separate metric names. For example,
memory usage:
node_memory_MemAvailable_bytes 1.00470784e+08
node_memory_MemFree_bytes 5.3211136e+07
node_memory_MemTotal_bytes 2.68435456e+08
Both are bounded (there's a fixed number of CPU modes, and a fixed number
of memory fields) so I'm not sure why the implementation differs between
the two.
Is this intentional? If so, is there a guideline or best practice as to
when to do the former vs when to do the latter?
Thanks!
--
You received this message because you are subscribed to the Google Groups
"Prometheus Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/prometheus-users/8722a500-e5c9-443e-8183-6bdb76699509%40googlegroups.com.
