Aman-Mittal commented on PR #625:
URL:
https://github.com/apache/fineract-backoffice-ui/pull/625#issuecomment-5849062732
`E2E (real Fineract, shard 2/3)` failed on `3df3800b` with one test:
`parity-screens.spec.ts:167` — "a pending loan is approved from the queue, in a
batch". Everything else on that commit is green, including all four mocked
shards, mobile, the other two real-Fineract shards, two-factor, unit tests and
TS/ESLint.
I do not believe it belongs to this branch, and here is the whole basis for
that rather than an assertion.
### It is not the translation gate
```
Error: expect(locator).toBeVisible() failed
Locator: getByTestId('queue-select-5')
> 196 | await
expect(page.getByTestId(`queue-select-${loan.loanId}`)).toBeVisible(...)
```
A plain visibility assertion. The batch approval itself succeeds —
`queue-result` appears and the loan leaves the approval queue, lines 189–193.
What fails is the loan not appearing in the **disbursal** queue afterwards.
`failOnMissingTranslations` is not in the call stack.
### What was ruled out
| Hypothesis | Result |
| --- | --- |
| This branch touches the screen | No. `work-queues.component.ts` is not in
the diff; 24 files changed, none in that flow |
| A missing `WORK_QUEUES.*` key | No. All 16 keys the screen uses are in
`en.json`; `i18n:check` reports 0 missing |
| Pre-existing on `main` | No. Every recent `e2e.yml` run on main is green,
including `a4770751`, this branch's base |
| `apache/fineract:latest` moved mid-session | No. Last pushed
`2026-09-25T21:25Z`, before every run today |
| Reproducible locally | No. `--project=backend e2e/parity-screens.spec.ts`
against a real Fineract: **9 passed**, including this test |
The commit it appeared on changes three files: a build-time script, one
`en.json` key, and one column label on the loan originators tab. I cannot
construct a path from any of those to the loan disbursal queue.
### What is odd, and worth someone else's eyes
It failed all three attempts with three freshly seeded loans (ids 5, 6, 7) —
deterministic within that run, not a race. That is why I am not calling it
flake and moving on. The difference I cannot test from here is that CI runs a
**fresh** Fineract while my local instance has hours of accumulated state; a
newly approved loan not appearing in the disbursal queue on a cold instance
would fit that shape.
A re-run of that shard would settle it. I do not have permissions to trigger
one.
### Separately
`deploy/docker-compose.yml` and `deploy/docker-compose-e2e.yml` both pin
`apache/fineract:latest`. It happens not to explain this failure, but it means
the backend under test can change with no commit in this repository, which
makes "green yesterday, red today, same code" unanswerable after the fact.
Filed as #628, independent of 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]
