Enhancing Observability in Rook Ceph
summary
Built end-to-end tracing for Ceph runs through Rook using OpenTelemetry instrumentation, a sidecar collector, and Jaeger storage and UI, with plans to replace the Jaeger agent by an OpenTelemetry collector. Traces link RGW calls through librados into OSD spans so bucket listing and S3 upload flows show where latency appears; a demo finds a misconfigured S3 path. Uses sampling to keep overhead low, deploys via Jaeger operator or Rook flags, tested with debug builds and Reef (Quincy) images as working proof.
AI-written summary · about
chapters (10)
- 0:00Introduction to talk
- 1:48Debugging challenges in Ceph
- 5:24Why tracing for Ceph
- 7:14OpenTelemetry and Jaeger
- 9:46How tracing works
- 15:03Tracing in Rook Ceph
- 18:26Demo of traces in Ceph
- 21:38Use cases for tracing
- 23:30Future of tracing in Rook
- 24:54Q&A session
related talks
22mReworking Observability In Ceph
41mRook - Running Ceph Using Kubernetes
26mRook: Intro And Deep Dive With Ceph Storage
31mRook: Intro and Deep Dive With Ceph Storage
55mIntroducing Rook
30mRook: Intro and Ceph Deep Dive
31mRook: Intro and Deep Dive With Ceph
29mSimplifying Storage in Kubernetes with Rook-Ceph