The GitHub Actions job "CI" on 
rocketmq-dashboard.git/fix/alert-dlq-resend-window has failed.
Run started by GitHub user unbridled-41 (triggered by unbridled-41).

Head commit for run:
9ba507e754d50877c62d077a4f59583f9d668287 / unbridled-41 
<[email protected]>
fix(alert): fire the DLQ resend alert on the documented 5-minute window

The bundled RocketMQDLQResendHigh rule announces

  "More than 10 dead-letter queue resends occurred in 5 minutes."

but thresholds with rate(), which is a per-second value:

  rate(rocketmq_dlq_resend_count[5m]) > 10

so the rule actually required ~10 resends per second (about 3000 per
5-minute window) before firing. A queue that starts re-driving dead letters
at a genuinely alarming rate — tens per minute — stays silent.

The sibling rule for the same "more than N in a window" contract,
rocketmq-consumer-rebalance.yaml, already uses
increase(rocketmq_consumer_rebalance_times[5m]) > 10, which is what the
annotation describes.

Fix: increase(rocketmq_dlq_resend_count[5m]) > 10, in the generator, the
shipped asset, and the dev-mode mirror.

Tested: 
AlertRuleAssetServiceTest#dlqResendAlertShouldCountResendsOverTheDocumentedWindow
(red on rate(), green after); the asset generator reproduces the checked-in
YAML byte for byte; checkstyle passes.

Report URL: 
https://github.com/apache/rocketmq-dashboard/actions/runs/37085849052

With regards,
GitHub Actions via GitBox

Reply via email to