From Billion...
summary
Claims a managed bring your own cloud observability backend that runs in the customer VPC, uses ClickHouse for log/tracing storage and VictoriaMetrics for PromQL-compatible metrics, and ingests hundreds of billions of rows per day while serving sub-second queries. Telemetry volumes double ~every 15 months and AI agents can produce ~50,000 spans in hours, so BYOC keeps data sovereign, avoids egress and vendor markups, tiers to object storage, and shows real incidents where full-fidelity data cuts troubleshooting from hours to minutes.
AI-written summary · about
chapters (8)
- 0:00Kubernetes monitoring intro
- 2:27Microservices observability challenges
- 9:49Where does all this telemetry go
- 13:30Self-hosted vs managed trade-offs
- 17:42Bring your own cloud for observability
- 21:11AI and external service observability
- 22:57Future of observability with BYOC
- 25:29Q&A
related talks
34mThe State of SQL-Based Observability
36mThe Power of Consolidation: A Unified Stack for Business Intelligence, Securi...
50mIntuit Journey To Unified Observability at Scale: Challenges, Be...
27mFast DIY O11y with OpenSrc Analytics & Visualization
34mReliable OpenTelemetry at Scale: No Queue, No Problem
26mO11y in One: Exploring ClickHouse as a Unified Telemetry Database
36m3 Pitfalls Everyone Should Avoid with Cloud Data
21mCloud-Native Observability for Financial Systems