How Simple Could Ceph Be?
summary
Ceph’s complexity traces to structure, function and operator cognitive load: the codebase exposes 2057 config options while only ~70 basic options are typically needed because most have defaults; cephadm provides a cephadm single-command deploy that bundles containers, Prometheus, logging and tracing; CLI already supports server-side completion and JSON/XML output but dry-run is spotty; proposals include friendly cluster names in metrics, service-discovery for OSDs, gRPC for transport, and replacing monitors with consul/etcd — ideas tested conceptually but not universally adopted.
AI-written summary · about
chapters (10)
- 0:00Introduction and motivation
- 1:47What does simple really mean
- 8:47Measuring simplicity in software
- 13:27CLI UX and improvements
- 18:42Simplifying deployment with cephadm
- 19:50Configuration complexity and defaults
- 24:09Reimagining Ceph's architecture
- 25:31Ceph as a platform vs simple storage
- 26:57Future ideas for operational simplicity
- 27:18Q&A: Trade-offs in simplicity
related talks
34mGetting Started as a Rook-Ceph Developer
06mLearn Ceph — For Fun, For Real, For Free!
37mCeph Manager Dashboard - The New Way To Manage Ceph & Gateway Management in Ceph Dashboard
43mBuilding the Production Ready EB level Storage Product from Ceph
37mA Glimpse of the New Ceph Messenger Built on Seastar
31mExtensibility without chaos: lessons from building Cozystack
26mFive Hundred Twenty-five Thousand Six Hundred K8s CLI’s
30mSturdy Clouds Creating Resilient Software Defined Data Centers