Alerting

Alerting helps you monitor cluster health and operational events by generating Alerts and delivering them to the people and systems responsible for responding. Alerts can be routed to external notification and event-management platforms, allowing you to incorporate ClusterWareAI ™ events into your existing operational workflows with systems such as email, Slack, PagerDuty, and other third-party tools. You can configure alerts with custom content using Mutators. The custom content can go to different target systems based on the severity, audience, or other requirements to make the notifications easier to understand and act upon.

Alerting in ClusterWareAI monitors configured event sources and generates alerts when matching events occur. Events are typically published through MQTT and may originate from ClusterWareAI services, the Cluster Health Monitoring System (CHMS), the Auto Remediation Service (ARS), or other ClusterWareAI components.

When an MQTT event matches a configured alert, the alerting system container reads the event data, processes the data with a mutator, assigns a severity level, and routes the resulting notification to one or more destinations. You can control which events generate alerts, how those alerts are presented, and where they are delivered. For example, you can configure an alert to go to a specific Slack channel if a compute node goes down and configure another alert to go to email when a new node moves from the Provisioning to Available state.