Monitoring at one of the largest retail groups in the world
summary
Large-scale retail monitoring runs on OMD with Naemon/Thruk for checks, Prometheus for metrics and Grafana as the single-pane view, covering 700,000 hosts across 53 zones and 226 locations with 280 backends and 5.7 million services. OMD sites run in parallel versions with add‑on RPMs, gearman scales check execution, LMD provides an in‑memory cache, exporters are auto‑discovered every 30 minutes, and a Playwright-based “space” pipeline deploys synthetic browser checks. A 16‑person core team handles ~800 tickets/month and leverages centralized CI for deployment and rollbacks.
AI-written summary · about
chapters (8)
- 0:00Introduction to monitoring
- 4:19Monitoring strategy overview
- 10:05OMD framework and tools
- 14:22Architecture and scaling
- 29:06End-to-end monitoring with Space
- 33:01Internal marketing and customer context
- 39:06Team and operations overview
- 41:20Q&A session
related talks
24mBanking Observability at Scale
23mMastering Monitoring & Alerting: Key to Seamless Performance
23mImplementation of Unified Observability at Scale From Scratch
56mGetting all the 99.99(9) you always wanted
52mHolistic Monitoring
42mEffective App Monitoring in Production
09mMonitoring? Not OPS problem
20mUsing Monitoring to Speed Up Your DevOps Journey