Driving O11y x to Nx: Probing Bullet Holes and not Breaking the Bank
summary
Argues that full‑stack observability must treat metrics, traces, logs and events the same, showing tool specifics: OpenTelemetry, Grafana/Prometheus, AWS X‑Ray/CloudWatch and New Relic; cites a customer with 260M monthly users generating 2.5PB of telemetry, ~3,000 engineers and four‑nines uptime. Warns DIY causes tool sprawl and hidden ops cost, commercial vendors give consolidation but unpredictable pricing; recommends DIY vs SaaS vs Cloud vs Hybrid tradeoffs and to instrument everything pre‑prod and prod while preferring consumption pricing and a free New Relic tier.
AI-written summary · about
chapters (13)
- 0:00Introduction and speaker background
- 1:44Bullet holes story and data lesson
- 4:32Observability patterns overview
- 7:22Scaling challenges for businesses
- 10:09Tool fragmentation and costs
- 13:59Support and documentation gaps
- 15:46Debugging production incidents
- 17:29Scaling and maintenance burdens
- 19:15Customer satisfaction at scale
- 21:00Instrumentation and data retention
- 22:23Granularity and data treatment
- 23:03Neuralik’s observability solution
- 25:08Pricing and free tier offer
related talks
38mThe Ticking Time Bomb of Observability Expectations
33mWhat Developers Should Know About Observability
39mThe Ticking Timebomb of Observability Expectations
32mEssential Software Observability for Every Role: Look No Further
56mFull-stack observability and application monitoring with
27mOpening the Door to Observability
30mBut Wait! There's...Still More‽ - Observability Data Volumes and Strategies for Managing Them
38mThe Murphy's Laws of Observability