The GitHub Actions job "CI" on rocketmq-dashboard.git/codex/instance-stale-rows-v2 has failed. Run started by GitHub user tju-yxq (triggered by tju-yxq).
Head commit for run: c766d30ca2e980a6517d8572965a5294119d11e6 / Yxq <[email protected]> fix(web): clear stale instance rows when a filtered reload fails The instance page's load failure path only shows a transient toast: after a type-filter or search change whose reload fails, the table keeps listing the rows fetched for the PREVIOUS query under the new filter — a DIRECT-only view silently showing PROXY rows, with nothing persistent indicating the data belongs to the old query once the toast auto-dismisses. The stale-response guard (request ids) already prevents racing responses from repopulating; the failure path of the latest request simply never cleared them. Mirror the sibling pages: BrokerCluster clears its data in the same catch, AlertRuleAssetList renders a persistent error instead. Clear the rows and the selection when the latest load fails, so the table can never show rows that contradict the active filter and no batch action can target rows that are no longer on screen. Report URL: https://github.com/apache/rocketmq-dashboard/actions/runs/38056113027 With regards, GitHub Actions via GitBox
