Is Building an Observability Platform with Open Source the Right Choice?
summary
Argues open source works if you accept integration and ops work: claims observability can eat ~30% of infrastructure cost, so using Prometheus, Thanos, Loki, Tempo plus OpenTelemetry and Perses saves licensing and enables customization. Demonstrates a homegrown correlator engine that links alerts, logs, metrics and traces across stores to cut troubleshooting time, and shows Prometheus recording rules producing 30/60/90‑day rightsizing recommendations and Kepler eBPF metrics for energy/sustainability. Recommends vendor neutrality and continuous evaluation.
AI-written summary · about
chapters (8)
- 0:00Introduction to observability
- 3:30Cost of ignoring observability
- 8:27Key features of solid platform
- 13:01Navigating open source landscape
- 17:14Right sizing with open source
- 19:18Visualization with Perses
- 23:58Why open source works
- 26:28Final takeaways and Q&A
related talks
45mOpen Source for Better Observability
28mOpen Source for Better Observability
42mOpen Source for Better Observability
12mOpen-source observability is the way forward, and here’s why
13mArchitecting an open observability stack
32mEssential Software Observability for Every Role: Look No Further
27mOpening the Door to Observability
41mNavigating the Open Source Observability Landscape