[TEST] Fix ZipkinExporterTestPeer.ExportJsonIntegrationTest stability

[tushar1977]

Fixes #3723

Changes

• Fix test ZipkinExporterTestPeer.ExportJsonIntegrationTest to expect 1 or 2 exports.
• This is needed when the batch processor is scheduled after the child span ends, but before the parent span ends.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 89.92%. Comparing base (16c7067) to head (2abfe1b).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3760      +/-   ##
==========================================
- Coverage   89.95%   89.92%   -0.02%     
==========================================
  Files         225      225              
  Lines        7158     7158              
==========================================
- Hits         6438     6436       -2     
- Misses        720      722       +2     

see 1 file with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[marcalff]

@tushar1977

Since you seem to be a student, allow me to provide some guidance on how to address this.

First, the original issue needs to be reproduced, to understand the problem.

I have done that, see the comments in #2723

Second, the fix needs to be tested locally to make sure it works, including when the original issue is seen.

With the following code:

  EXPECT_CALL(*mock_http_client, Post(expected_url, _, IsValidMessage(report_trace_id), _, _))

      .Times(Between(1, 2)) /* <-- proposed fix */
      .WillOnce(Return(ByMove(ext::http::client::Result{
          std::unique_ptr<ext::http::client::Response>{new ext::http::client::curl::Response()},
          ext::http::client::SessionState::Response})));

  child_span->End();
  std::this_thread::sleep_for(std::chrono::milliseconds(1000)); /* <-- how to reproduce the bug */
  parent_span->End();

the test still fails, so the fix is incorrect.

[malff@malff-desktop build-otelcpp-abiv1]$ ./exporters/zipkin/zipkin_exporter_test 
Running main() from /builddir/build/BUILD/googletest-release-1.11.0/googletest/src/gtest_main.cc
[==========] Running 4 tests from 1 test suite.
[----------] Global test environment set-up.
[----------] 4 tests from ZipkinExporterTestPeer
[ RUN      ] ZipkinExporterTestPeer.ExportJsonIntegrationTest

GMOCK WARNING:
/home/malff/CODE/MARC_GITHUB/opentelemetry-cpp/exporters/zipkin/test/zipkin_exporter_test.cc:159: Actions ran out in EXPECT_CALL(*mock_http_client, Post(expected_url, _, IsValidMessage(report_trace_id), _, _))...
Called 2 times, but only 1 WillOnce() is specified - returning default value.
Stack trace:
terminate called after throwing an instance of 'std::runtime_error'
  what():  
    The mock function has no default action set, and its return type has no default value set.
Aborted (core dumped)

There is more to it.

Last, once you have a fix that works locally when the bug is exposed, you can verify the fix still works in the normal case, by removing the added sleep_for.

Only then will be patch be ready.

What needs to be resolved at this point is Called 2 times, but only 1 WillOnce() is specified

[tushar1977]

Hi @marcalff
I have fixed and checked this issue in my local (by including the sleep and removing it). Previously, the issue was that i was doing 2 number of times but i was calling WillOnce only once.
So there were 2 approaches to solve this. One was to include WillOnce twice and other was using WillRepeatedly. But the functionByMove will be performed only once. To tackle that, i used Invoke method which creates a fresh object on every call.

[tushar1977]

Its failing on macOS and maintainer mode. Let me check that

[marcalff]

LGTM, thanks for the test fix.

[marcalff]

@tushar1977

See edits to the title, and PR comments.

[tushar1977]

@tushar1977

See edits to the title, and PR comments.

sure i'll make sure of it in future PR's