Observability

Effective observability is crucial for managing dynamic preview environments. Diverge is designed to be highly observable, providing operators with clear visibility into the system’s state.

Core Principles

1. Native Metrics: Diverge exposes metrics in the Prometheus exposition format.
2. Controller Visibility: Standard controller-runtime metrics are available for the operator.
3. Health Checking: Granular endpoints for liveness, readiness, and controller status.

Component Monitoring

• Controller: Monitors API interactions, reconcile durations, and queue depth.
• Server: Monitors RPC requests, latency, and error rates.
• Proxy: Tracks routing decisions and connection states.

These metrics enable teams to build comprehensive Grafana dashboards and set up reliable alerting rules.