composer/WARN/2023_005
Product: Cloud Composer
Rule class: WARN - Something that is possibly wrong
Description
Cloud Composer runs a liveness DAG named airflow_monitoring, which runs on a schedule and reports environment health. If the liveness DAG run finishes successfully, the health status is True, which means healthy. Otherwise, the health status is False. Note that the environment health could be intermittently unhealthy due to events like scheduled maintenances. However, overall it should be healthy.
Remediation
The environment health metric depends on a Composer-managed DAG named airflow_monitoring that is triggered periodically by the airflow-monitoring pod. If the DAG isn’t deleted and doesn’t report failures in the Airflow UI, check the airflow-monitoring logs to see if there are any problems related to reading the DAG’s run statuses.
Otherwise, you can check if there was a maintenance operation aligned with the unhealthy time frame from the monitoring dashboard. Note that maintenance operations happen during the configured maintenance window. If your current maintenance window does not fit to your business requirements, please consider changing it.