Beyond Particle Physics: The Impact of Ceph and OpenStack on CERN's Multi-D...
summary
CERN runs 100 PB Ceph for block, object and CephFS and an OpenStack cloud that exposes bare metal, VMs and virtual networking across more than 450k CPUs; the estate includes ~11,000 servers, ~0.7 EB on tape and a new off‑campus computing center linked by a 10 Tb line. The team deploys RGW multisite with the secondary set to read‑only (manual failover), experiments with CephFS stretch clusters and hits operational pain points: MDS metadata imbalance, RGW sync hotspots, sparse multi‑site monitoring.
AI-written summary · about
chapters (7)
- 0:00Introduction and CERN overview
- 8:29OpenStack and Ceph integration
- 13:28New data center construction
- 20:53Ceph metadata scaling challenges
- 30:26Ceph RGW multisite setup
- 35:02Ceph stretch clusters exploration
- 37:32Closing remarks and hiring
related talks
15mThe CERN Cloud Infrastructure: 5 years in 15 minutes
21mKeynote: CERN Experiences - Ricardo Rocha & Clenimar Filemon
26mPetabyte Scale Object Storage Service Using Ceph in A Private Cloud
21mCeph for Public Cloud Workloads
30mDeploying and Operating CephFS for Scientific Applications at Fermilab
06mPractices of Ceph Object Storage in Public Cloud Services
16mA novel Cloud High Performance Computing infrastructure for user
33mEnhancing Ceph for Infinite Scalability at LINE