731 B
731 B
Architecture Update: Alerting & Anti-Spam
Context from User (Hendrik): "I would expect we also need to add some kind of monitoring / alerting to our platform. using your example, we need to detect the failure at 3 am and be able to notify but not spam an operator."
Directives:
- Anti-Spam / Grouping: We need smart alerting (grouping by route/business entity, debouncing, threshold-based). 100 failed messages = 1 alert, not 100 emails.
- Local Evaluation: Since TSDB/Logs are on the Customer Hub, alert evaluation MUST happen on the Hub (e.g., using
vmalertandAlertmanager). - SaaS Management: The SaaS Control Plane UI is where users configure the rules. These rules are pushed down to the Hub.