monitoring-expert

To set up Prometheus metrics and Grafana dashboards, you instrument backend services to expose metrics, configure Prometheus to scrape them, and build Grafana views for latency, error rate, throughput, and business KPIs.

The best way to design structured logging for production troubleshooting is creating JSON logging pipelines that include request IDs and apply safe field redaction to protect sensitive data.

To trace a failing dependency in a distributed system, you wire in OpenTelemetry tracing instrumentation to correlate events across distributed services, isolating the bottleneck.

Yes, you can forecast capacity for growth by profiling CPU and memory bottlenecks, running load tests, and analyzing performance signals to predict future resource requirements.

You create actionable alerts for error spikes by defining Prometheus alert rules that trigger based on specific metric thresholds, ensuring teams are notified of production issues.

You need OpenTelemetry to implement distributed tracing and correlate events across services, while health checks can be configured independently to monitor service availability.