action #167272
closedci: test coverage in t/05-scheduler-full.t less than 100% size:S
Description
Observation¶
Seen in
https://app.codecov.io/gh/os-autoinst/openQA/pull/5910/indirect-changes
Allegedly the subroutine scheduler_step
is not covered:
note "No worker with ID $id active"; # uncoverable statement
}
sub scheduler_step { OpenQA::Scheduler::Model::Jobs->singleton->schedule() }
my $worker_settings = [$api_key, $api_secret, "http://localhost:$mojoport"];
Looking at the Devel::Cover HTML report in the artifacts
https://output.circle-artifacts.com/output/job/c7f4dc57-bc3f-4b5d-b590-060b89c7e75e/artifacts/0/cover_db/t-05-scheduler-full-t.html
line 100 has a zero, although we see in line 106 and others, that this subroutine is executed.
Locally I cannot reproduce it.
Acceptance criteria¶
- AC1: codecov checks in pull requests again all pass including t/
Suggestions¶
- Try to reproduce locally with coverage analyis enabled, e.g.
make coverage KEEP_DB=1 FULLSTACK=1 TESTS=t/05-scheduler-full.t
- If not reproducible easily just do code analysis and try out in PRs triggering circleCI+codecov checks online
Updated by okurz about 2 months ago
- Subject changed from ci: test coverage in t/05-scheduler-full.t less than 100% to ci: test coverage in t/05-scheduler-full.t less than 100% size:S
- Status changed from New to Workable
Updated by mkittler about 2 months ago
- Status changed from Workable to In Progress
- Assignee set to mkittler
Updated by mkittler about 2 months ago
The problem is also visible on codecov showing exactly one missed line under t
: https://app.codecov.io/gh/os-autoinst/openQA/pull/5943/tree (The 100 % are confusing, though - probably a rounding issue.)
(The file on codecov: https://app.codecov.io/gh/os-autoinst/openQA/pull/5943/blob/t/05-scheduler-full.t#L100)
Updated by mkittler about 2 months ago
PR (only workaround): https://github.com/os-autoinst/openQA/pull/5950
Updated by openqa_review about 2 months ago
- Due date set to 2024-10-11
Setting due date based on mean cycle time of SUSE QE Tools
Updated by mkittler about 2 months ago
- Status changed from In Progress to Feedback
- Priority changed from Urgent to High
We're using the same version of perl-Devel-Cover
on the CI then I used when I tried to reproduce the problem under Tumbleweed. Maybe it is the different version of Perl itself that makes the difference but we probably don't want to change that right now so I am not investigating this further.
This PR will hopefully fix the issue without having to mark the line as uncoverable: https://github.com/os-autoinst/openQA/pull/5955
Updated by okurz about 2 months ago
- Due date deleted (
2024-10-11) - Status changed from Feedback to Resolved
PR merged. And https://app.codecov.io/gh/os-autoinst/openQA/tree/master/t shows 100% so good again.