Issue 7475 - Fix CI test failures

[vashirov]

Description:
retrocl, resource_limits and clu test suites are flaky and fail from time to time.

Fixes: #7475

Summary by Sourcery

Stabilize flaky CI tests in retro changelog, export task timeout, and file-descriptor limit suites.

Bug Fixes:

• Adjust retro changelog trimming test flow to start without trimming and apply aggressive trimming only after generating entries to avoid timing-related flakiness.
• Relax export task timeout expectations so the test tolerates tasks that complete before the initial timeout instead of failing.
• Use fixed default max descriptor limits in file-descriptor limit tests to avoid environment-dependent failures when system limits vary.

[sourcery-ai]

Hey - I've found 1 issue, and left some high level feedback:

• In fdlimits_test.py, hardcoding SLAPD_DEFAULT_MAXDESCRIPTORS = 1048576 may make the test brittle if the default changes or differs across environments; consider deriving this value from the running server (e.g., reading the current nsslapd-maxdescriptors) instead of embedding a magic number.
• In test_task_timeout for the export task, when the task finishes before the timeout you now only log and skip the timeout path; consider explicitly marking this as a skipped/xfail scenario (e.g., via pytest.skip or a separate test) so the test intent is clearer and the behavior shows up explicitly in test reporting.

Prompt for AI Agents

Please address the comments from this code review:

## Overall Comments
- In `fdlimits_test.py`, hardcoding `SLAPD_DEFAULT_MAXDESCRIPTORS = 1048576` may make the test brittle if the default changes or differs across environments; consider deriving this value from the running server (e.g., reading the current `nsslapd-maxdescriptors`) instead of embedding a magic number.
- In `test_task_timeout` for the export task, when the task finishes before the timeout you now only log and skip the timeout path; consider explicitly marking this as a skipped/xfail scenario (e.g., via `pytest.skip` or a separate test) so the test intent is clearer and the behavior shows up explicitly in test reporting.

## Individual Comments

### Comment 1
<location path="dirsrvtests/tests/suites/clu/dsconf_tasks_test.py" line_range="264-268" />
<code_context>
     task = ExportTask(inst)
     task.export_suffix_to_ldif(export_ldif, DEFAULT_SUFFIX)
     task.wait(timeout=.5, sleep_interval=.5)
-    assert task.get_exit_code() is None
-    task.wait(timeout=0)
+    if task.get_exit_code() is None:
+        task.wait(timeout=0)
+    else:
+        log.info('Export task completed before timeout - skipping timeout check')

     # Test timeout for schema validate task
</code_context>
<issue_to_address>
**issue (testing):** The timeout test can now pass without actually asserting timeout behavior if the export completes early

This weakens the test’s value and can hide regressions in the timeout path. To keep it non-flaky but still meaningful, you could:
- Skip the test when the task finishes early (e.g. `pytest.skip("export finished before timeout")`) instead of just logging; or
- Make the export reliably slow enough to exceed the timeout and retain the assertion that `get_exit_code()` is `None` before the second `wait`; or
- Split into separate tests for timeout behavior and successful completion.

Any of these would ensure the timeout logic is still explicitly verified.
</issue_to_address>

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook

_{Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.}

[sourcery-ai]

issue (testing): The timeout test can now pass without actually asserting timeout behavior if the export completes early

This weakens the test’s value and can hide regressions in the timeout path. To keep it non-flaky but still meaningful, you could:

• Skip the test when the task finishes early (e.g. pytest.skip("export finished before timeout")) instead of just logging; or
• Make the export reliably slow enough to exceed the timeout and retain the assertion that get_exit_code() is None before the second wait; or
• Split into separate tests for timeout behavior and successful completion.

Any of these would ensure the timeout logic is still explicitly verified.

[packit-as-a-service]

Congratulations! One of the builds has completed. 🍾

You can install the built RPMs by following these steps:

• sudo dnf install -y 'dnf*-command(copr)'
• dnf copr enable packit/389ds-389-ds-base-7476
• And now you can install the packages.

Please note that the RPMs should be used only in a testing environment.

[progier389]

Looks good but some minor copmments

[progier389]

LGTM