Advanced Monitoring
Beyond the standard Prometheus/Grafana/AlertManager setup covered in the Monitoring Guide, KubeOpera runs three dedicated backend services specifically to unify observability signals that don't fit neatly into a single metrics pipeline — worth knowing about if you're trying to trace a request, tail logs, or track SLO compliance programmatically rather than through the dashboard.
tracing-gateway fronts distributed tracing, reading from Jaeger's query API rather than a raw OTLP collector — so if you're instrumenting a service to emit traces, point it at your existing Jaeger collector as normal, and tracing-gateway is what the dashboard and MCP Server tools query afterward. log-gateway does the equivalent for logs. slo-manager and apm-gateway track error-budget burn and application performance metrics respectively, each with its own dedicated dashboard page (see Dashboards Reference).
All four are reachable both through the dashboard and directly as MCP Server tools (get_traces, get_logs, get_slo_status, get_apm_metrics) — useful if you want to pull this data into a script or an AI agent's context rather than reading it off a page.