Monitoring

Kovra provides built-in monitoring powered by VictoriaMetrics and Grafana. Each cluster gets automatic metric collection and dashboards.

Cluster Monitoring

Navigate to Monitoring to see cluster-level metrics:

• CPU Usage — Cluster-wide and per-node CPU utilization
• Memory Usage — RAM consumption across nodes
• Pod Status — Running, pending, and failed pods
• Node Health — Node conditions and resource pressure

Application Monitoring

Each application has its own monitoring dashboard showing:

• Request Rate — HTTP requests per second
• Error Rate — 4xx and 5xx response rates
• Latency — P50, P95, P99 response times
• Resource Usage — CPU and memory per pod

Configuring Monitoring

Monitoring is enabled per-cluster:

1. Go to Clusters → select your cluster
2. Click Configure Monitoring
3. Monitoring agents will be deployed automatically
4. Metrics start flowing within minutes

Alerting

Set up alerts to get notified when metrics cross thresholds:

1. Navigate to your application
2. Click Alerts tab
3. Choose from preset alert rules or create custom ones
4. Configure notification channels (Slack, email, webhooks)

See Applications for more details on per-application alerting.