Prometheus Learning Roadmap
Goal: From no monitoring to production observability with Prometheus + Grafana
Learning Phases
Phase 1: Basics (1 week)
•What Prometheus is: pull-based time series
•Install with Docker or Kubernetes operator
•prometheus.yml: scrape configs
•Explore metrics in Prometheus UI
Phase 2: PromQL (2 weeks)
•Instant vectors and range vectors
•Aggregation: sum, avg, rate, irate
•Filtering with label matchers
•Functions: increase, topk, histogram_quantile
Phase 3: Alerting (1 week)
•Recording rules for pre-computed queries
•Alert rules: expr, for, labels, annotations
•Alertmanager: routing, receivers, silences
•Integrations: Slack, PagerDuty, email
Phase 4: Integration (2 weeks)
•Node Exporter for host metrics
•kube-state-metrics for Kubernetes
•Application instrumentation (client libraries)
•Service discovery: kubernetes_sd_configs
Phase 5: Scale & Production (2 weeks)
•Federation for hierarchical Prometheus
•Remote write to Thanos/Mimir/Cortex
•Long-term storage with Thanos
•Grafana dashboards from PromQL
Job Roles This Enables
•SRE
•DevOps Engineer
•Platform Engineer
•Observability Engineer
Target Certifications
•Grafana Associate covers Prometheus + Grafana
How to Use This Roadmap
1.Work through phases in order — each builds on the previous
2.Complete the labs section for each phase before moving on
3.Build a project after each phase to cement learning
4.Use the interview section to test understanding regularly
5.Track progress: revisit earlier sections as concepts compound
Prerequisites
See prerequisites.md in this section for what you should know before starting.