Conversation
|
|
||
| meter | ||
| .gaugeBuilder("process.runtime.jvm.system.cpu.load_1m") | ||
| .setDescription("Average CPU load of the whole system for the last minute") |
There was a problem hiding this comment.
This definition from the semantic convention doesn't really match the the description from the OperatingSystemMXBean. It's a pretty odd calculation as it isn't bounded to the range [0,1], [0, 100] or even to [0, 100*N] where N is the number of processors.
This stack overflow post describes it as the "average is the number of waiting threads", and says the upper limit is equal to the number of threads you have.
There was a problem hiding this comment.
Isn't this the same as unix load average that is displayed for example by uptime?
...ntime-metrics/library/src/main/java/io/opentelemetry/instrumentation/runtimemetrics/Cpu.java
Outdated
Show resolved
Hide resolved
breedx-splk
left a comment
breedx-splk
left a comment
There was a problem hiding this comment.
The casting bit to work around multiple implementations is a drag. Bummer there isn't a common interface. Otherwise, looks like a nice straightforward impl. Thanks!
.../java/io/opentelemetry/instrumentation/javaagent/runtimemetrics/RuntimeMetricsInstaller.java
Show resolved
Hide resolved
...ntime-metrics/library/src/main/java/io/opentelemetry/instrumentation/runtimemetrics/Cpu.java
Outdated
Show resolved
Hide resolved
...ntime-metrics/library/src/main/java/io/opentelemetry/instrumentation/runtimemetrics/Cpu.java
Outdated
Show resolved
Hide resolved
...ntime-metrics/library/src/main/java/io/opentelemetry/instrumentation/runtimemetrics/Cpu.java
Outdated
Show resolved
Hide resolved
4 participants
Reflects the semantic conventions in #2292.