GBIM adds Prometheus, Grafana, and k6 for end-to-end observability

GBIM has rolled out an observability stack using Prometheus, Grafana, k6, and correlation IDs to track system behavior from frontend to backend [devto]. The setup delivers real-time metrics, performance telemetry, and traceability for debugging.

Custom Prometheus metrics now capture business-critical events: gbm_auth_register_total{role,outcome}, gbm_auth_activation_total{outcome}, and gbm_admin_account_verification_total{action,outcome} among others. These track registration, activation, and verification outcomes by role and status, linking technical performance to business flow.

k6 smoke tests run continuously, generating load and capturing response times, error rates, and throughput [devto]. The tests feed into Grafana dashboards, where engineers monitor system health in real time. Frontend analytics events—such as user registration and activation—are also logged, closing the loop between user action and backend processing.

Correlation IDs are injected at the API gateway and propagated across services, enabling engineers to trace a single request through multiple systems. This reduces mean time to debug by allowing teams to isolate failures in distributed flows, especially during authentication and admin verification.

The metrics expose bottlenecks—for example, email send latency tracked via gbm_auth_email_send_duration_seconds{event,outcome}—and help validate service level objectives. With Grafana dashboards now reflecting both technical and business KPIs, teams can correlate spikes in error rates with drops in activation rates.

One gap remains: the k6 tests currently run in staging only, limiting production insight. Extending them to production-like environments would strengthen reliability feedback.