Opentracing shim#1909
Conversation
Codecov Report
Additional details and impacted files@@ Coverage Diff @@
## main #1909 +/- ##
==========================================
- Coverage 87.32% 87.29% -0.02%
==========================================
Files 166 166
Lines 4673 4673
==========================================
- Hits 4080 4079 -1
- Misses 593 594 +1
|
|
Is it better to move this into a standalone directory? |
|
Thanks for the PR. Tagging @rnburn as someone contributed to both otel-cpp and opentracing-cpp if he would like to review this. though I don't see him active on Github for long now, just trying the luck :) |
|
The copyright check in CI is new, please fix: |
marcalff
left a comment
There was a problem hiding this comment.
In third_party/prometheus-cpp, the git submodule for prometheus changed.
It looks like a merge issue with:
commit 9200d0c8fb4684e5fffd2c6915c8117a747c8e2b
Author: Lalit Kumar Bhasin <lalit_fin@yahoo.com>
Date: Sat Feb 4 20:32:33 2023 -0800
upgrade prometheus-cpp to v1.1.0 (#1954)
Please fix.
marcalff
left a comment
There was a problem hiding this comment.
If I understand correctly, the Opentracing shim introduces a dependency on the opentracing-cpp repository, which is a deprecated library, per:
It is a bit unusual to add dependencies like that, but it makes sense in this case, so ok.
Not knowing enough of Opentracing, it is a bit hard to review the code for correctness for every details.
That said, the overall code looks ok.
The contributed code is protected by a WITH_OPENTRACING option, and is separate from the opentelemetry-cpp code which is otherwise unchanged, so there is very little technical risks with this PR, in my opinion.
SpanContextShim::extractFrom() has a hack with kUniqueTag, but after looking at it I don't have a better proposal to offer, so ok with it.
Please fix the prometheus-cpp spurious change, and I will approve this PR.
Great work.
marcalff
left a comment
There was a problem hiding this comment.
LGTM, thanks for the contribution.
lalitb
left a comment
There was a problem hiding this comment.
Sorry for responding late on it. Changes in general look good, and clean approach - the shim confined to separate directory, and behind feature flag.
Most of the existing contributors doesn't have expertise on Opentracing, and bandwidth to manage this component, so @chusitoo it would be appreciated if you can own any related issues/changes going ahead :)
@lalitb Thanks for the review. Not a problem, I am not an expert in OpenTracing myself but I can support any related issues or requests related to the shim. |
Fixes #78, answers to #1180
Changes
For significant contributions please make sure you have completed the following items:
CHANGELOG.mdupdated for non-trivial changes