Cephalocon 2019

Cloud Service Providers (CSPs) constantly face pressure to keep up with the processing demands of the big data workloads. The data footprints for the big data workloads span several petabytes which means hundreds of terabytes of new storage infrastructure acquisition every year. At such a scale, any inefficiencies in capital expenditures (Capex) and resource allocation are greatly magnified. High capacity NVMe QLC NAND flash delivers balanced TCO with improved rack density while addressing the processing demands of big data workloads.

Running analytics frameworks using Ceph backed data lake storage provides flexibility to scale compute and storage infrastructure independently.
 
In this session, we will outline Ceph all flash data lake architecture, share analytics workload performance benchmark results (e.g., TPC-DS), provide best practices and outline Ceph community work.