You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Some of the issues raised in spec and semconv are related to interaction between application and instrumentation code (or different layers of instrumentations):
classifying errors - it's context-dependent
enriching telemetry - e.g. add specific attributes to metric depending on the specific context call is made in.
suppressing all nested instrumentations or duplicates in certain scope
I agree that this is a big area that will need multiple sub-issues, but if we're going to work on any of them, we should keep the wider area in mind.
We should avoid designing APIs for each individual problem that might not work or look ugly together.
So I think we need to have a general idea on how to solve the big problem and then we can break it down into smaller pieces that play nicely with each other.
Some of the issues raised in spec and semconv are related to interaction between application and instrumentation code (or different layers of instrumentations):
See open-telemetry/opentelemetry-specification#4131 (comment) for the context.
Such customizations are sometimes possible for spans (using SpanProcessors), but not in general possible for metrics.
I'd like to propose tackling this issue in scope of the devex project.
The text was updated successfully, but these errors were encountered: