Alerting Operations

Typical operational checks

• verify that jobs.alerts is enabled and populated for the tenant
• confirm message, flow, or keystore source jobs are producing current runtime data
• compare alert spikes with worker timing and recent configuration changes
• validate whether repeated alerts are new events or continued open conditions

Common failure patterns

• alert definitions exist, but source data is stale
• alerts appear missing because related sync jobs are disabled
• acknowledgement changes state in the UI, but the underlying condition remains active
• old generic alert expectations are confused with the current CPI alert engine