Career Track · OSS-First · Hands-on 80 / Theory 20

100 Days of Open-Source Observability

A deep-expertise curriculum for an experienced Staff engineer — not a beginner tour. Eleven phases take you from the golden signals to operating the full LGTM + OTel + eBPF stack, debugging injected production incidents with telemetry alone, and contributing to Grafana / OpenTelemetry OSS. One GitHub repo per lab, four certifications, one production capstone.

100 days · ≈14 weeks Target window: Jun 22 → Sep 29 2026 (shift to your start date) Roles: Grafana · Chronosphere · Honeycomb · Datadog · Platform Eng Assumes: Node.js · AWS · Kafka · K8s fundamentals
11
Phases
~55
Daily repos
4
Certifications
4
Telemetry signals
1
Production capstone

01 The 100-Day Timeline

Time runs left→right in days (week ticks). Click any phase bar to jump to its full plan below. Certifications and capstone milestones sit on their own lanes.

↔ Scroll horizontally on narrow screens. Diamonds = certification exams · orange markers = capstone build milestones.

02 Phases & Daily Labs

Each phase lists the concepts, the per-day GitHub repos, the milestone gate you must clear before moving on, the certification it feeds, and concrete OSS contribution targets. Click a header to expand.

03 The System You'll Operate

The demo system is built incrementally across all phases — by day 100 every hop emits metrics, logs, traces, and profiles, with eBPF as a zero-code cross-check.

Observability stack (all OSS)

04 Certification Plan

Pursue in priority order — each exam is scheduled right after the phase that prepares you, so you certify on fresh hands-on reps.

PriorityCertExam ~DayPrepared byWhy it matters for your target roles

05 Capstone — day100-production-capstone

Deploy the full system across four targets, then break it six ways and prove you can diagnose each failure using telemetry only. This repo is the centerpiece of your portfolio.

Deploy across four targets (progressive)

① Docker Compose — single-host full stack ② Kind — local K8s, Operator-driven scrape ③ Kubernetes — Helm LGTM + OTel Operator injection ④ EKS — managed prod, IRSA, remote-write to Mimir, real load

Inject failures · debug with telemetry only

🎯
Capstone deliverables: a README portfolio index linking all ~55 repos · an architecture diagram · SLOs + burn-rate alerts + a runbook per alert · a recorded "incident debugged via telemetry only" walkthrough · a short blog post / write-up per game-day. This is what you show in a Grafana / Honeycomb interview.

06 Training Sources & OSS On-Ramps

Reference material — kept to 20% of your time. The repos below are your standing OSS contribution targets.

Courses & talks

    Standing OSS contribution targets

      07 How to Run This (for an experienced engineer)