[rhoai-2.25] RHAIENG-6036: fix trustyai papermill test CSV paths

[jiridanek]

Summary

Cherry-pick the trustyai fairness-test data fix from ODH main (#3070, f5327ea6c): use local income-unbiased.csv / income-biased.csv instead of stale ubi9-python-3.11 raw GitHub URLs that 404 in CI.

Also drops the duplicate read_csv in test_fairnessmetrics (already gone on main tip via opendatahub-io#3363).

Full cherry-pick -x f5327ea6c does not apply — that commit is a large 3.4-ea2 relock; this PR carries only the notebook hunk.

Test plan

• make test-jupyter-trustyai-ubi9-python-3.12 papermill: test_fairnessmetrics + test_datafairness pass

Related

• Split out of [rhoai-2.25] RHAIENG-6036: extend OCP port-forward HTTP wait for slow arches #2488
• Upstream: Update Notebooks to use new 3.4-ea2 artifacts (RHAB Images & RH-Index)  opendatahub-io/notebooks#3070

Made with Cursor

[coderabbitai]

Warning

Review limit reached

@ide-developer, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 38 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 51f7d5a5-4d09-49dc-8938-205ba4f8c94b

📥 Commits

Reviewing files that changed from the base of the PR and between 7344630 and 9f37bd1.

📒 Files selected for processing (1)

• jupyter/trustyai/ubi9-python-3.12/test/test_notebook.ipynb

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/rhoai-2.25-trustyai-local-csvs

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[ide-developer]

@coderabbitai full review

[coderabbitai]

✅ Action performed

Full review finished.

---

Your plan includes PR reviews subject to rate limits. More reviews will be available in 38 minutes.

[openshift-ci]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: ide-developer
Once this PR has been reviewed and has the lgtm label, please assign caponetto for approval. For more information see the Code Review Process.

The full list of commands accepted by this bot can be found here.

Details

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[jiridanek]

CI failure analysis (run 29087867712)

This run did not fail on the CSV-path fix. Build + testcontainers pytest (16 passed) succeeded; failure is in the k8s papermill phase (make_test.py).

What failed (step 32, ~line 346)

test_jupyter_with_papermill.sh aborts with exit 1 after:

error: unable to upgrade connection: container not found ("notebook")
make: *** [Makefile:319: test-jupyter-trustyai-ubi9-python-3.12] Error 1

The pod was never stable enough to kubectl exec / kubectl cp:

1. make_test.py wait_for_stability — pod never became Ready (~3 min of timed out waiting for the condition)
2. _wait_for_workload in papermill script — another 600s wait, also timed out
3. Script continued anyway → exec/cp against a crashlooping pod → hard fail

Root cause (captured in dumped k8s diagnostics)

The finally block dumps show the real issue — liveness probe kills the container before Elyra/Jupyter listens on 8888:

From kubectl get events / kubectl describe pods:

Warning  Unhealthy  Liveness probe failed: dial tcp …:8888: connect: connection refused
Warning  Unhealthy  Readiness probe failed: Get "http://…:8888/notebook/opendatahub/jovyan/api": connection refused
Normal   Killing    Container notebook failed liveness probe, will be restarted
Warning  BackOff    Back-off restarting failed container notebook

Probe config still at defaults: liveness initialDelaySeconds: 5, readiness 10. TrustyAI/Elyra needs ~12s+ to start. Pod ended 0/1 Ready, 9 restarts.

The papermill fairness tests (income-*.csv) were never exercised in this run.

Fix path

This is the probe-timing issue addressed in #2487 (initialDelaySeconds: 15 on liveness). Suggest merging #2487 before this PR (or cherry-picking the statefulset probe hunk here), then re-run CI.

Locally validated on k3s with Quay image + this branch: papermill passes once probe delay is patched (same class of failure as CI).