aminghadersohi commented on PR #44287:
URL: https://github.com/apache/superset/pull/44287#issuecomment-5687454163
Final CI audit at head `48931bcf0423997558f70e795d91da5c2dc92726` — all
checks terminal, none pending/cancelled/action-required: **66 pass, 12 skipped
(conditional), 15 fail**. Workflow-level: 17 success, 3 skipped, 5 failure; no
zero-job runs.
All 15 failures were inspected individually. 12 execution jobs
(`test-sqlite`, `test-mysql`, `test-postgres (current)`, `test-postgres-hive`,
`test-postgres-presto`, `test-load-examples`, `docker-build (dev)`, 4
playwright jobs, `cypress-matrix (chrome)`) stop at the same base-branch error:
```
ERROR [flask_migrate] Error: Multiple head revisions are present for given
argument 'head'
```
The remaining 3 (`test-postgres-required`, `playwright-tests-required`,
`cypress-matrix-required`) are aggregate gates propagating that result. **Zero
failures are attributable to this PR's diff.** `unit-tests (current)`,
`pre-commit (current)`, `docker-build (lean)`, `docker-build (superset)` and
all 20 testcontainers jobs pass.
Independently recomputed from `origin/master` @
`a184395eca7c291d0b8b438f00690c633112f4d7`: 387 revisions, **2 heads** —
`c7f53d184ea2` and `88a01c781622`. Master is broken for every PR, not just this
one. Note that the `Enforce single Alembic migration head` check reports
success here only because it short-circuits when a PR touches no
`superset/migrations/` files; it is not evidence of a single head.
No rebase performed: this PR is `MERGEABLE` with no conflicts, and master
still carries both heads, so a rebase cannot change any outcome.
**Coordination hazard worth flagging before either lands:** #44288
(`e2f3a1b9c640`) and #44283 (`f7b3a9c14e02`) *both* declare `down_revision =
("88a01c781622", "c7f53d184ea2")`. They merge the same two heads. If both land
as authored, master ends up with two heads again and this exact blocker
returns. Whichever merges second must be rebased to chain off the first as a
single parent rather than re-merge the same pair.
**Exact human action required (maintainer-only):**
1. A Superset committer reviews and merges #44288 — a no-op merge revision
joining exactly the two heads above; it is fully green (65 pass / 16 skipped /
0 fail).
2. Before merging #44283, rebase its `down_revision` onto whichever merge
revision lands first.
3. Once master is single-head, this PR needs a re-run to clear the 15 red
jobs, plus a committer approval (`reviewDecision: REVIEW_REQUIRED`).
Dependency claim re-verified directly against PyPI metadata: pinotdb
`6.0.0`/`7.0.0` declare `sqlalchemy (>=1.4,<2)` under `extra == "sqlalchemy"`;
`8.0.0` is the first release declaring `sqlalchemy (>=2.0,<3)`, unchanged
through `9.1.2`. The `pinotdb[sqlalchemy]>=8.0.0,<10.0.0` floor and the
matching UPDATING.md / metadata / docs wording are correct.
Known limitation, stated plainly:
`test_dialect_supports_installed_sqlalchemy` uses
`pytest.importorskip("pinotdb.sqlalchemy")` and `pinotdb` is not in
`requirements/development.txt`, so that contract assertion skips rather than
executes in `unit-tests (current)`. This follows the existing repo pattern
(`tests/unit_tests/db_engine_specs/test_datastore.py`); adding the driver to
dev requirements is deliberately out of scope here.
Nothing merged, nothing drafted, no migration workaround folded into this PR.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
