Open adispennette opened 11 months ago
We do need for someone do look under the hood and write proper docs for metrics.
Is pretty much reporting Timers
for http client execution, encoding/deconding and overall time from invokation of feign method.
So you can track if a request is slow, and whos to blame
Thanks for the comment @velo I am attempting to monitor the connection pools for each connection. I think I have the right metric with feign.Client.count
or feign.Feign.count
since they seem to be the same thing. I am also looking at injecting a custom okhttp client that has the pool monitor configured in as well. But I do like the other metrics.
the count
will only diverge when there are retries. 99% of times it will be the same.
In addition it'd be helpful to document metrics published by MicrometerObservationCapability
.
In addition it'd be helpful to document metrics published by
MicrometerObservationCapability
.
That would be awesome. Please send a PR, I will make sure to approve it.
is there any documentation on what the various metric published by micrometer mean? I am making a lot of guesses but don't actually know for sure what they are relaying.