The GitHub Actions job "CI" on rocketmq-dashboard.git/fix/audit-cleanup-truncation has failed. Run started by GitHub user unbridled-41 (triggered by unbridled-41).
Head commit for run: 82f19d775a81a4400cb7aaa00bd3c630975d41f5 / unbridled-41 <[email protected]> fix(audit): disclose a cleanup that hit its per-call batch ceiling POST /audit-logs/cleanup deletes at most CLEANUP_BATCH_SIZE x CLEANUP_MAX_BATCHES (500 x 20 = 10000) rows and answered with the count alone: return auditRepository.deleteBefore(cutoff, CLEANUP_BATCH_SIZE, CLEANUP_MAX_BATCHES); and the page did not even read the count: await cleanupAuditLogs(cleanupDays); message.success(t('audit.cleanupSuccess', { n: cleanupDays })); So a deployment with more than 10000 expired rows was told the retention window was applied while older rows survived - the operator had to notice a stale row in the refreshed list to learn otherwise, and the audit record of the cleanup (added for the deletion itself) says the same "deleted=10000, SUCCESS". Report whether the sweep hit its ceiling (a full sweep means expired rows may remain), thread it through the endpoint response and the web client, and warn with the real count instead of claiming success; a non-truncated cleanup keeps the existing message. Tests: - server: cd server && mvn -o -B -ntp test -Dtest='AuditServiceTest,AuditControllerTest' -> Tests run: 32, Failures: 0, Errors: 0. New cleanupLogsReportsWhenTheSweepHitItsCeilingTest pins the signal (and the existing bounded-batch case now asserts truncated=false); three controller cases were updated to the new service return type. - web: cd web && npx vitest run src/pages/ops/__tests__/AuditPage.test.tsx -> Tests 12 passed. New case "discloses a cleanup that hit the per-call cap" fails before this change (the success toast is shown) and passes after it. - npx eslint reports no errors; npx tsc -b passes; checkstyle passes. Report URL: https://github.com/apache/rocketmq-dashboard/actions/runs/37991031576 With regards, GitHub Actions via GitBox
