suxiaogang223 commented on PR #69886:
URL: https://github.com/apache/doris/pull/69886#issuecomment-6997674045
Local self-review: **PASS** for `c7ce034535b3ebf8a638f4d3f6715e6c711e6bd7`
against `29c1083f65fe92bb5a778a660fde3e8671ef537b` after correcting the
External regression assertion.
External build 6908 failed because it expected the old SDK message, while
the actual SQL error was `Paimon Rust estimated input row size 67 exceeds
writer budget 1`. The test now expects `exceeds writer budget 1`. The
underlying writer correctly rejected the row.
Validation: recorded-error matching and two unrelated-error negative
controls passed; `run-regression-test.sh --run -d
external_table_p0/paimon/write -s test_paimon_rust_write_append -dryRun`
selected one suite with no script errors. This checks script selection/parsing,
not SQL execution. The local Doris/Spark/MinIO services are stopped; full
execution awaits the External rerun.
FE UT build 6871 is unrelated to this PR: 23,116 passed, 77 skipped, zero
failures/errors. Its collector reads `Running
...InnerProductTest2026-11-18T...` as a nonexistent class ending in `2026` and
reports missing XML, even though the real class has three passes. No common CI
helper is changed.
<details>
<summary>Doris code-review Part 1.3 conclusions</summary>
| Checkpoint | Conclusion |
| --- | --- |
| 1. Goal and proof | Fix the single failing External suite. With a one-byte
budget, the INSERT correctly fails before Arrow allocation. The updated
expectation matches that budget-specific error. |
| 2. Scope and reuse | One assertion string changes; the SQL input, one-byte
limit, snapshot checks, production code and output fixtures are unchanged. |
| 3. Concurrency | No concurrency code changes. The prior per-task
writer/admission/report ownership conclusions remain applicable. |
| 4. Lifecycle | No lifecycle or static initialization changes. Writer
credit, file ownership and failure cleanup paths are unchanged. |
| 5. Configuration | The existing temporary one-byte configuration and
restoration remain intact. No configuration/default changes. |
| 6. Compatibility | No API, runtime or format changes; the accepted
compatibility scope from the prior review is unchanged. |
| 7. Thrift/Protobuf | No protocol diff from the prior reviewed head/base;
no IDs, enum values or optionality change. |
| 8. Parallel paths | Rust/JNI selection and all read/write implementations
are unchanged. Only the Rust row-budget error expectation is corrected. |
| 9. Conditions/errors | The assertion remains specific to exceeding writer
budget 1. It does not accept unrelated SQL/network errors or suppress a
successful INSERT. |
| 10. Coverage | The same External suite still executes the failure and
verifies the committed snapshot afterward. Existing BE UT covers a row
exceeding budget before Arrow allocation. |
| 11. Test results | Recorded CI failure: old expected text does not match;
corrected text does; connection/missing-table negative controls do not. Preset
regression runner dry-run selected and parsed one suite without script errors.
Full SQL execution awaits External CI; no full local pass is claimed. No .out
changes. |
| 12. Observability | No profile, log or metric changes. |
| 13. Transactions/persistence | No transaction, commit, recovery or EditLog
changes; snapshot verification remains in the suite. |
| 14. Atomicity/cleanup | The observed INSERT is rejected before Arrow
allocation and commit; the existing post-failure row/snapshot checks are
retained. |
| 15. FE/BE propagation | No new variable, descriptor, switch or
commit-message change. |
| 16. Performance/memory | No production performance or allocation changes;
the test still enforces the same one-byte admission budget. |
| 17. Other issues | FE UT build 6871 has zero test failures/errors and
Maven BUILD SUCCESS. The common collector misparses interleaved Running/log
timestamps as InnerProductTest2026; the real InnerProductTest has three passes.
This unrelated helper is left unchanged as requested. |
</details>
<!-- doris-repo-review:v1:begin -->
```yaml
schema: doris-repo-review/v1
status: PASS
pr: apache/doris#69886
commit: c7ce034535b3ebf8a638f4d3f6715e6c711e6bd7
base: 29c1083f65fe92bb5a778a660fde3e8671ef537b
reviewed_at: 2026-11-19T02:18:49+00:00
reviewer: suxiaogang223
model: gpt-6.1-sol
effort: xhigh
findings: {blocker: 0, major: 0, minor: 0, nit: 0}
rounds: 1
converged: true
```
<!-- doris-repo-review:v1:end -->
--
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]