The GitHub Actions job "CI" on rocketmq-dashboard.git/codex/stale-rows-global has failed. Run started by GitHub user tju-yxq (triggered by tju-yxq).
Head commit for run: c6c9d76c2beee3706fce7b8fd396f0a905dc6288 / Yxq <[email protected]> fix(web): clear stale rows on every page when a filtered reload fails A failed list reload after a filter, search, scope, or pagination change left the PREVIOUS query's rows listed under the new filter on ten pages, behind either a transient toast (eight pages) or a persistent error banner (two pages). The instance page already cleared its rows in the same situation; this extends that discipline everywhere the pattern existed: - consumer groups: old instance's groups stayed when the search/scope reload failed - topic list: stale topics under a new type/search filter (the client-side visibleTopics mask did not cover type changes or pagination) - audit feed: stale records and total under new filters - notification deliveries: stale feed under new channel/status filters - system alerts: stale alerts under new level/domain filters - alert rules: stale rules and their selection under new search/enabled filters - data sources and cloud credentials: stale lists under new search/type/vendor filters - DLQ groups and detail messages: the error banner explained the failure but the rows beneath it kept listing the previous search/page - message query results: the query error banner sat above the previous query's rows as if they answered the new inputs Every catch now clears the rows and the total (and the selection where one exists) alongside the existing error feedback, so the table can never show rows that contradict the active filter. Racing responses remain guarded by the existing request-id/generation checks. Report URL: https://github.com/apache/rocketmq-dashboard/actions/runs/38063430738 With regards, GitHub Actions via GitBox
