oteps icon indicating copy to clipboard operation
oteps copied to clipboard

Span Status Improvements (without errors)

Open mwear opened this issue 4 years ago • 4 comments

We recently discussed OTEP #123 at the error working group and agreed it introduced some good ideas, but they were coupled with error reporting. Exception data was handled by https://github.com/open-telemetry/opentelemetry-specification/pull/697 and the successHint inspired this error.hint PR: https://github.com/open-telemetry/opentelemetry-specification/pull/822.

There are currently proposals to remove Span.status (see: https://github.com/open-telemetry/opentelemetry-specification/issues/706). If we decide to keep it, this adaption of OTEP-123 could be an improvement worth considering.

See the original OTEP for commentary and rationale.

mwear avatar Aug 18 '20 02:08 mwear

There are currently proposals to remove Span.status (see: open-telemetry/opentelemetry-specification#706). If we decide to keep it, this adaption of OTEP-123 could be an improvement worth considering.

What is the Error Working Group position on this? Do we remove Status? Do we keep it but improve? Do we have 2 opposing proposals because Error Working Group was not able to come to a concensus on this?

I would like to have more clarity on this otherwise I am not sure which of the opposing proposals to actively pursue. The OTEP to remove Status is also submitted: https://github.com/open-telemetry/oteps/pull/134

Just to be clear: I am happy if the goal is to discuss 2 proposals in parallel and chose a winning one, just want to understand if that is the intent here.

tigrannajaryan avatar Aug 19 '20 14:08 tigrannajaryan

We had much more agreement on removing current status code, than on its replacement. So we want to remove Google RPC code bases status before GA, that is what my otep is about. And separately we discuss what should we put instead of it. Either before GA or after. And for that there are several options. One is this PR. Another is open-telemetry/opentelemetry-specification#822

iNikem avatar Aug 19 '20 15:08 iNikem

I think we need to require having a reasonable replacement for conveying erroneous state in the Span. I believe this is needed before GA. Ideally we will first know what is this replacement before removing the Status. I am also happy if the answer is that we don't need any replacement, but I would like that analysis to also be done before removal of the Status.

tigrannajaryan avatar Aug 19 '20 16:08 tigrannajaryan

@mwear Should this be closed in favor of #136 ?

iNikem avatar Sep 17 '20 18:09 iNikem

@mwear please reopen if you change your mind about this

tedsuo avatar Feb 06 '23 17:02 tedsuo