How to monitor your business, and not just your pods
summary
Argues for monitoring business KPIs end-to-end instead of per-pod, demonstrates building a four-step delivery funnel dashboard that sums M1 rates across pods and uses a custom active delivery carriers metric, a successful delivery API calls metric and a conversion metric (delivery completions divided by contact-detail events). Uses Loki, Prometheus, Grafana for logs, metrics and visualization, breaks down results by labels per carrier, warns about label-cardinality explosions and shows alerting to on‑call channels; cites Wix scale of 150 million metrics/hour and half a trillion logs/day.
AI-written summary · about
chapters (9)
- 0:00Introduction to Business Driven Monitoring
- 1:22The Importance of Monitoring Business KPIs
- 3:17Tools for Monitoring: Loki, Prometheus, and Grafana
- 4:38Implementing Business Driven Monitoring
- 7:38Creating Visualizations in Grafana
- 8:30Real-World Example: Wix Delivery System
- 11:50Demo: Building a Business Funnel Dashboard
- 21:51Enhancing Dashboards with Alerts and Investigations
- 23:14Conclusion and Final Thoughts
related talks
34mMetrics driven development
28mImprove Monitoring and Observability for Kubernetes with OSS Tools
24mPractical Kubernetes Monitoring with Prometheus
14mFrom Metrics to Mastery: Monitoring vital signs of Kubernetes
29mKubernetes-Defined Monitoring [I]
56mFull-stack observability and application monitoring with
41mMetrics Applied
48mMetrics All the Way: Data Driven Operations