Lightning Talk: Faster Insights and Improved Accuracy: Spotify's Prometheus Upgrade
summary
Spotify replaces Heroic with VictoriaMetrics and Prometheus, completing a two year migration that yields 10x faster queries and large cost savings. Migration includes rewriting hundreds of HQL queries to PromQL, updating metric agents, switching alerts to VMAlert and Alert Plumber, and building Backstage migration trackers plus query converters. Results show 90% adoption of the new monitoring path, faster Grafana panels, deprecated Heroic datasets, and more accurate, real‑time alerts.
AI-written summary · about
chapters (8)
- 0:00Introduction to monitoring
- 1:21Why replace Heroic?
- 2:54Selection and overhaul
- 4:15Managed Monitoring (MMA)
- 5:39Migration challenges
- 7:17Results and impact
- 8:46Future plans and takeaways
- 11:03Conclusion
related talks
05mLightning Talk: How the Observability Team at Spotify Radically Decreased On-Call Alerts
31mObservability: Dust Off Prometheus with VictoriaMetrics
01h 03mWhat’s new in the Prometheus ecosystem?
28mObservability Project Updates
13mLightning Talk: A Drop-in System To Accelerate Metrics Observability by 100x Us...
28mHow to reduce expenses on monitoring with VictoriaMetrics
31mEverything You Need to Know About Prometheus 3.0
40mScale Your Monitoring Solution With the VictoriaMetrics Ecosystem