Simplified self-diagnostics example #2274
Conversation
|
@lalitb We may actually be able to get-rid of the dedicated self-diag example, and instead demonstrate this in every examples as well.. Maybe something to iterate for next release, based on feedback, especially since we have now starting doing extensive internal logging. |
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #2274 +/- ##
=====================================
Coverage 79.5% 79.5%
=====================================
Files 121 121
Lines 20960 20960
=====================================
+ Hits 16671 16673 +2
+ Misses 4289 4287 -2 ☔ View full report in Codecov by Sentry. |
| .add_directive("tonic=error".parse().unwrap()) | ||
| .add_directive("h2=error".parse().unwrap()) | ||
| .add_directive("tower=error".parse().unwrap()) | ||
| .add_directive("reqwest=error".parse().unwrap()); |
There was a problem hiding this comment.
we can remove this filter if none of these transport dependencies are used in the example.
There was a problem hiding this comment.
let me move it to the otlp examples, so when users face this issue, they have something to easily refer to, to mitigate the infinite logs problem.
There was a problem hiding this comment.
will add this shortly in a follow up PR.
Simplifies self-diag example to use stdout exporter, and trigger internal log by using an invalid metric instrument name. The purpose is to demonstrate internal logging, so this seems the easiest setup to trigger one.
I kept the minimum filter example to show how to prevent the infinite, recursive logging problem. More complex examples might be needed in practice, but I think it is better instead of focus on fixing the root cause as priority than workarounds which are hacky. (But they are a necessary hack now)