Observability & Reliability | Prometheus + Grafana by Mircea StefanescuObservability & Reliability | Prometheus + Grafana by Mircea Stefanescu

Observability & Reliability | Prometheus + Grafana

Portfolio project
Mircea Stefanescu
This personal project turns a local application's health signals into a practical monitoring and incident-diagnosis workflow. The Reliability Pack includes Grafana dashboards for request rate, error rate, p95 latency, CPU and memory, together with Prometheus alert rules and incident runbooks.
The implementation combines Prometheus, Grafana, OpenTelemetry, Loki and Mimir. A Python receiver accepts Slack and PagerDuty protocol payloads locally, validates them, records alert events in SQLite and deduplicates repeated submissions. Controlled delivery failures make retry and recovery behavior inspectable without contacting real messaging or incident-management accounts.
The repository includes configuration, receiver tests, setup instructions and captures from the local demonstration on 16 October 2026. The dashboard image shows example service metrics; additional evidence and verification steps are linked in the repository.
This is a personal demonstration on one host, with synthetic telemetry where documented. Production availability, live New Relic integration and customer outcomes have not been established.
For a remote freelance project, share your stack and the monitoring problem you want to investigate so we can scope a reproducible first milestone.

0 likes · 0 views · Nov 16, 2026

Personal observability demo with Prometheus, Grafana dashboards, alert rules and a local paging receiver. Source code, tests and dated local evidence.

Tags