Prometheus Intro and Deep Dive
summary
Prometheus overview plus recent changes: claims of handling 10 million active time series and ~1 million samples/sec with about 1–2 bytes per sample compression, default two-week retention but no hard upper limit, and built-in scaling via Thanos/Cortex/Mimir. New PromQL features include negative offsets, the @ modifier, trig functions and present_over_time. Feature flags control experimental changes; agent mode (no local storage) and a remote_write receiver (push ingestion) arrive alongside build-time plugin modularization and upcoming sparse high‑resolution histograms.
AI-written summary · about
chapters (15)
- 0:00Intro
- 1:25What is Prometheus
- 2:06Architecture
- 7:13Selling Points
- 7:33Data Model
- 12:54Querying
- 16:52Alerting
- 17:23Efficiency
- 19:04Bridging the gap
- 20:08Conclusion
- 22:15New in v2.25: Feature flags
- 24:06What's new in PromQL
- 26:15New in v2.32: Agent mode
- 27:31New in v2.25: Remote write receiver
- 30:54What's coming
