Event Details

For the past 5 years, CERN IT has employed Ceph technology to build scale-out storage for our OpenStack cloud. For the block and object storage use-cases, with and without erasure coding, Ceph has demonstrated to be flexible and scalable while being resiliant to infrastructure failures.

But as our computing requirements evolve, the importance of a scale-out HPC filesystem is emerging. We therefore present usage of CephFS -- managed by Manila -- in production for our HPC and general IT infrastructure. We highlight the key metrics required by our users, including POSIX compliance, small-file latency, metadata throughput and scalability, and fault tolerance, while showing the results of industry-standard and new microbenchmarks.