The GitHub Actions job "CI" on rocketmq-dashboard.git/fix/ai-group-detail-delay-unknown has failed. Run started by GitHub user Frun1na (triggered by Frun1na).
Head commit for run: feaeeb9c34765162a0a9d6b22fc7810a8e06ab51 / Apulupie <[email protected]> fix(ai): stop publishing an unmeasurable consume delay as zero seconds The rmq.group.detail tool output published delaySeconds at its zero default when the provider could not measure a delay, so the model read "the group is caught up" from data the same response marks unavailable: the provider leaves delaySeconds at 0 and records the reason in consumptionTimestampAvailable (RocketMQMetadataProvider keeps zeros when the broker reports no consumed-message timestamp), and the console-side fix for the same root cause is #5920/#5919. The tool contract dropped the flag and published the number, while its sibling signals already carry their unknown states (onlineInstances = -1, Progress.totalLag = -1, retryMaxTimes omitted). delaySeconds is now an Integer that is null - and therefore omitted by the record's NON_NULL inclusion - when consumptionTimestampAvailable is false, so an unmeasurable delay is absent instead of reported as a measured zero. group.yaml drops delaySeconds from the output's required list and documents the absent case; catalog_gen.go is regenerated from the shard. A delay that was measured is unchanged. Verified with `mvn -B -ntp test -Dtest=ConsumerGroupReadToolHandlersTest` (10 tests green, 0 checkstyle violations), `mvn -B -ntp test -Dtest=ToolOutputSchemaContractTest` (2 green) and `make catalog-verify` (green). Report URL: https://github.com/apache/rocketmq-dashboard/actions/runs/38039417685 With regards, GitHub Actions via GitBox
