What you can observe about Che

Che exposes metrics, logs, and telemetry data that help you monitor platform health, diagnose issues, and track workspace usage.

Che provides three categories of observability data:

Metrics

The Che server exposes JVM metrics (memory usage, classloading, garbage collection) on port 8087 via the che-host Service /metrics endpoint. The DevWorkspace Operator exposes workspace startup, failure, and performance metrics on port 8443 via the devworkspace-controller-metrics Service /metrics endpoint. The Kubernetes in-cluster monitoring stack (Prometheus) scrapes both endpoints. The Che Operator automatically creates the required ServiceMonitor, Role, and RoleBinding resources. You can view these metrics on custom dashboards in the Kubernetes web console.

Logs

The Che server uses configurable log levels for individual loggers. You can increase verbosity for specific components to isolate issues, log HTTP traffic between Che and the Kubernetes API server, and collect diagnostic logs with the chectl command-line tool. For log configuration procedures, see How server logging works.

Telemetry

Workspace activity events (start, stop, editor actions) can be sent to Segment and Woopra by using the built-in Woopra telemetry plugin, or to a custom analytics backend by using a telemetry plugin you build yourself. For details on building a custom plugin, see Creating a telemetry plugin.