Farm Convergence with DRBD
summary
Claims a working, cheaper alternative to Ceph RBD by wiring DRBD replicas into chains of mirrors and exposing them as convergent storage; uses DRBD v9 dis feature and dual‑primary RAID1‑over‑network behavior, places volumes on top of LVM, and runs LVM thin provisioning snapshots as writable VM images. Scripts implement a simple three-node PoC with static allocation, live migration support, no split‑brains over two years of production use, and “good enough” latency versus GFS2.
AI-written summary · about
chapters (7)
- 0:00Introduction and background
- 3:15DRBD basics and principles
- 8:16DRBD architecture with LVM
- 10:05Snapshots as active volumes
- 12:57Live demo setup
- 23:05Storage orchestration concepts
- 27:44Guest orchestration and Q&A
related talks
34mModernizing Ceph Deployments at CERN: CephFS and Object Storage Across Data Centres
49mIntro to Ceph, the Distributed Storage System
02h 59mCeph Storage: a Caffeinated Primer
45mBuilding Available and Partition Tolerant...
20mRemote Replication in MicroCeph: RBD and Beyond
21mNew Techniques to Improve Small I/O Workloads in Distributed File Systems
16mFORD: Fast One-sided RDMA-based Distributed Transactions for Disaggregated Persistent...
26mImplementing a Resilient Ceph RGW Object Storage with Erasure Coded Data Pool as the Foundation...