pilichoumao opened a new pull request, #6433: URL: https://github.com/apache/rocketmq-dashboard/pull/6433
### Which Issue(s) This PR Fixes - Fixes #6423 ### Brief Description For group totalLag=-1 and queue lags [17,-1], Consumer health diagnostics show both the useful known subtotal 17 and a reported total 17. All-unknown or absent observations become a reported zero. Keep the reported total unavailable unless there is a separately measured aggregate or a nonempty queue result with every lag available. Keep the subtotal and use its positive evidence for the existing offline-group warning. The existing page already renders null as unavailable; no UI text, API, collection, threshold or scoring changes. #5926 separately handles unreadable consume stats and retains its availability guard when composed with this fix; #5564 handles diagnostic translations. This patch does not include either contributor's implementation. README documents the distinction. ### How Did You Test This Change? Node 22.19.0 / npm 11.11.1 and unchanged lockfile. Tests use `NODE_OPTIONS=--max-old-space-size=768 UV_THREADPOOL_SIZE=1 GOMAXPROCS=1 npm test -- --maxWorkers=1 --pool=forks --no-file-parallelism`; all stages ran serially. Lint/build use 1024 MiB old-space. - Diagnostic + ConsumerPage full files: baseline **5 failed / 62 passed**, fixed **67/67 passed**. Includes mixed/all-unknown/no observations, known aggregates including zero, fully measured queue totals, the actual health tab, and preservation of positive-backlog offline evidence. One old assertion expecting zero for all-unknown queues was corrected. Page services are mocked; no live Broker test is claimed. - Temporary combination with #5926's actual five-file patch and its tests: **69/69 passed**. The three overlapping utility sites were explicitly composed; this is behavior compatibility, not a claim of conflict-free Git merging. Nothing from that patch is included in the submitted diff. - Full frontend: **1611 passed / 2 failed**, 159 files. Unchanged failures are Producer's “discards a slow connection response after the topic changes” (`pointer-events: none`) and UserManagement's “exports all users that match the active filters” (missing `search: ops`). A clean archive of this exact 779b8c0 baseline previously reproduced both complete files as **29 passed / 2 identical failures**, with the same lockfile and single-worker options. Full tests are **not all green**. - `npm run lint`: 0 errors / 10 existing warnings. `npm run license:test`: 5/5. `npm run build`: TypeScript, Vite and distribution-license check passed. Modified-file Prettier and `git diff --check` passed. - Backend unchanged; JVM/ArchUnit and real Broker tests not rerun. ### Checklist - [x] One coherent change; unrelated modifications are not bundled in - [x] Conventional commit - [x] Regression tests with failing baseline and positive controls - [x] No new UI strings - [x] ASF headers retained; no dependencies added - [x] Documentation updated -- 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]
