Set 14.2.0 as upcoming version; run N-1 tests against 14.1.0

[ihalatci]

Validation experiment — NOT a version bump

This PR consumes the-process #228's dev images (image suffix -dev-9490d30, with /opt/citus-versions/v14.1.0 baked in) to validate whether the latent #8587 N-1 gap reproduces on the 14.x line, mirroring citus #8632 on the 13.x line.

Part B only — .github/workflows/build_and_test.yml (10 lines):

• image_suffix: -v4271d84 → -dev-9490d30
• All 4 N-1 jobs: citus_libdir → /opt/citus-versions/v14.1.0, citus_libdir_label → v14.1.0, citus_version → 14.1-1

No changes to configure.ac, control files, or SQL stubs.

Expected result (prediction)

The N-1 slot on the 14.x line is PG18. Prediction:

Job	n_1_mode	Prediction
Lib N-1 / PG18 check-multi-1	all	FAIL
Coordinator N-1 / PG18 check-multi-1	coordinatoronly	FAIL
Worker N-1 / PG18	workeronly	PASS
SQL N-1 / PG18	(SQL only)	PASS

Why — the 2 reds are pre-existing / orthogonal to the image bump

Regress test create_ref_dist_from_citus_local, line SELECT citus_internal.delete_placement_metadata(1);:

• The expected .out encodes post-Enforce object ownership for more citus-internal UDFs #8587 behavior (could not find any shard placement with id 1).
• The N-1 image .so (both v14.0.1 and v14.1.0) is pre-Enforce object ownership for more citus-internal UDFs #8587 and emits This is an internal Citus function can only be used in a distributed transaction.

Only coordinator-side loads of the N-1 .so diverge, so Lib N-1 and Coordinator N-1 fail while Worker N-1 and SQL N-1 pass. Because the error string is identical for v14.0.1 and v14.1.0, the image bump does not cause the failure — it is a latent #8587 gap, the same pre-existing condition accepted on 13.x in #8632.

Do not merge

Observe-only. This PR exists to run the N-1 jobs (which only run on pull_request) and confirm the pass/fail matrix above.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 88.62%. Comparing base (1c80ba2) to head (a7a79b4).

Additional details and impacted files

@@               Coverage Diff                @@
##           release-14.0    #8633      +/-   ##
================================================
- Coverage         88.63%   88.62%   -0.01%     
================================================
  Files               288      288              
  Lines             64116    64115       -1     
  Branches           8046     8045       -1     
================================================
- Hits              56827    56820       -7     
- Misses             4969     4971       +2     
- Partials           2320     2324       +4     

🚀 New features to boost your workflow:

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

[ihalatci]

Closing this observe-only experiment PR. Its purpose — validating that the latent #8587 N-1 check-multi-1 gap (citus #8632 on the 13.x line) reproduces on the 14.x line — is complete and CI-confirmed across two runs (dev images -dev-9490d30 and stable -vf4adc47): Lib N-1 / PG18 + Coordinator N-1 / PG18 FAIL on the #8587 gap while Worker N-1 + SQL N-1 PASS, exactly as predicted. This branch carries only workflow-only experiment commits. The real Set 14.2.0 as upcoming version work is being prepared as a fresh, self-contained branch off release-14.0 (mirroring the 13.4.0 template commit 4cd5cdd), which supersedes this PR. Not merging.