unbridled-41 commented on PR #3346: URL: https://github.com/apache/rocketmq-dashboard/pull/3346#issuecomment-5570682434
Thanks @lizhimins — both blocking points are reworked in 86483815: 1. **Genuine 5.5.0 evidence captured.** Reproduced against a live 5.5.0 name server (docker `apache/rocketmq:5.5.0`) querying a route-less topic through this repo's `rocketmq-tools` 5.5.0 client. The client throws `MQClientException` with `responseCode=17` and `errorMessage="No topic route info in name server for the topic: no-such-topic"` — confirming the message-substring branch keyed on a string the client never produces (dead code), and the previously quoted red-test output did not reflect the 5.5.0 client. The PR description now carries the genuine capture and the stale quote is flagged as replaced. 2. **Grading moved to the shared response-code classifier.** The hand-rolled code check and the message-substring fallback are deleted; both catch blocks now grade via `MqResponseCodes.hasResponseCode(e, ResponseCode.TOPIC_NOT_EXIST)` inside the generic `catch (Exception e)`, matching the convention from #3302 (`RocketMQMessageProvider`'s TOPIC_NOT_EXIST/NO_MESSAGE grading). A failure carrying route-absent-looking text with a different response code surfaces as 502 — pinned by new `getTopicRoutesGradesByResponseCodeOnly` / `getTopicConsumersGradesByResponseCodeOnly` tests; the empty-result tests now mock the genuine 5.5.0 exception shape. Tests: `RocketMQMetadataProviderTest` 40/40; `MetadataServiceTest` 36/36, `TopicControllerTest` 16/16, `MqResponseCodesTest` 6/6; full suite 2094 tests with only the 3 pre-existing baseline failures (re-verified identical on pristine `origin/rocketmq-studio` f5519e87) — zero new failures. -- 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]
