Bringing a Ceph Based Enterprise Email System Into the Field
summary
Deutsche Telekom migrates its email storage to Ceph using the open source librmb library that writes each message as a RADOS object and stores mutable attributes in omap; the cluster uses SSD-hosted metadata and HDD object storage with erasure coding. A live field test moves 1.1M accounts (~440M messages, ~400 TB user data, ~4B EC chunks) and runs for 6 months field test with no customer data loss. Key findings: slow MDS failover and hot‑standby complexity, MDS pinning and index bottlenecks, dovecot plugin fixes required, ~2.5‑hour recovery after a rack outage and ~7‑hour recovery when filling to capacity.
AI-written summary · about
chapters (7)
- 0:00Introduction and system overview
- 10:30Technical requirements and setup
- 18:06Phase one load and failure testing
- 21:16Phase two real customer migration
- 28:01Findings and cost analysis
- 31:30Future plans and conclusions
- 34:28Q&A session
related talks
26mPetabyte Scale Object Storage Service Using Ceph in A Private Cloud
33mEnhancing Ceph for Infinite Scalability at LINE
17mNetGSM's Storage Journey: Modernization with Ceph
30mDeploying and Operating CephFS for Scientific Applications at Fermilab
06mPractices of Ceph Object Storage in Public Cloud Services
21mCeph for Public Cloud Workloads
05mThe Enterprise Ceph Spectrum
28m10 Billion Messages in a Day, How We Built the Infrastructure That Delivers