Log Analytics: Understanding Complex Systems at Scale
summary
Humio is a streaming query engine plus storage that indexes only metadata and uses compression, per-chunk bloom filters and a time index to search historical logs interactively; an early customer pushes 1M events/second (20 TB/day) from 30,000 desktops. The stack runs on-prem or cloud, re-architected cloud‑native for production, serves ~150 paying customers including a Bloomberg integration, and proves you can trade indexing CPU/disk for fast, parallel compressed scans that store up to 100x more data and power dashboards, alerts and ad hoc incident digs.
AI-written summary · about
chapters (6)
- 0:00Intro
- 0:55How did I end up here?
- 14:27Demo: What is Humio?
- 17:27How does it work?
- 31:35Some lessons learned
- 38:00Outro
related talks
40mObservability that scales: Analyze data in stream and query remotely
46mOperational Logs and Event Management for HPC S...
16mLog Rehydration and Logs to Metrics
28mFrom Fragmented Logs to Unified Insights: Scaling O11y
29mWide Event Analytics
43mBeyond Logs: Real-Time Anomaly Detection with Machine Learning
30mData @Scale Seattle- Sundaram Narayanan
42mMake your logs work for you