Monitoring the deployment

What to watch so that a problem in the platform is noticed before a customer notices it.

In the console

  • Overview › Collector health shows every source with its last event time and lag. A stalled source raises a notification.
  • Settings › System shows database health, size, growth, version and session limits.
  • Notifications of kind source stalled and containment tampered are the ones to route to an on-call channel.

On the host

  • docker compose ps: every service up and healthy.
  • The health endpoint on the gateway hostname, which also reports the number of connected agents.
  • Disk usage on the data volume. Set an alert in your infrastructure monitoring at 80 percent.

Exporting metrics

XenXDR exposes metrics in Prometheus format for your existing monitoring. The deployment files list the endpoints.