Ceph Telemetry - The Why, What, and How
summary
Ceph Telemetry collects anonymized cluster data from 3,400 clusters totaling about 1.7 exabyte and ~200,000 OSDs, split into five channels: basic, crash, device, ident, perf. Reporting is opt-in (ceph telemetry on, ceph telemetry show, diff, collection ls) and strips FSIDs, host and pool names, and serials. Crashes ship backtraces to a crash processor that deduplicates signatures and a Redmine bot that files or links issues, letting developers track regressions and feature adoption quantitatively.
AI-written summary · about
chapters (10)
- 0:00Introduction to Ceph Telemetry
- 1:04Why Ceph Telemetry matters
- 3:57What is Ceph Telemetry
- 7:56How Ceph Telemetry works
- 11:36Public Telemetry dashboards
- 15:20Observability into features
- 19:01Deprecating unused features
- 20:07Crash reporting and analysis
- 23:46Crash dashboard insights
- 29:28Call to action and Q&A