Group all cpu usage per core statistics under cpus #496
Conversation
| System *bool | ||
| Proc *bool | ||
| Filesystem *bool | ||
| Cpu_per_core *bool |
There was a problem hiding this comment.
Can you use here `CpuPerCore *bool yaml:"cpu_per_core" to make sure we get the right config? We should also add it for the above 3 configs as we did in the other beats to make it clear, the values are coming from a yaml doc.
In the example above the backticks are missing, as they are also use to mark some code :-( here is an example: https://github.com/elastic/filebeat/blob/master/config/config.go#L39
There was a problem hiding this comment.
Yes, I agree with you it's better to add "yaml". I didn't reach the point to make nicer the code.
There was a problem hiding this comment.
@monicasarbu Sorry, was too early. No review label yet ;-)
monicasarbu
force-pushed
the
group_cpu_usage_per_core
branch
2 times, most recently
from
2f3bb9e
to
3c4c394
Compare
monicasarbu
force-pushed
the
group_cpu_usage_per_core
branch
from
3e7a2c9
to
a536471
Compare
|
This feature is implemented only for Unix systems, so I am disabling the test for windows in order for the tests to pass on windows. |
Add a configuration option to enable adding cpu usage per core. By default is false as there are too many data exported. Rename proc option with process (as the type it got changed to process in the last release). Add an example JSON output Add a test for cpu_per_core option only on Unix systems
monicasarbu
force-pushed
the
group_cpu_usage_per_core
branch
from
b5f365f
to
a14ecfd
Compare
|
Ready for review. |
|
|
||
| === cpui Fields | ||
|
|
||
| This group contains cpu usage statistics of the core i, where i is the core number |
There was a problem hiding this comment.
I think it would look better to use cpuX with X being the variable rather then "i".
Group all cpu usage per core statistics under cpus
The goal of this PR is to group all the cpu usage per core statistics under
cpusand add a configuration optioncpu_per_coreto enable it. By default it is set to false as it exports a lot of data.The number of fields that are exports depends on the number of cores, so for these data we are not providing sample dashboards as it depends on the number of cores.
This is the first change out of multiple changes requested by #424.