Seppo Sakari Heikkila (CERN)
Cloud storage is an emerging architecture aiming to provide increased scalability and access performance, compared to more traditional solutions. CERN is evaluating this promise using Huawei UDS and OpenStack storage deployments, focusing on the needs of high-energy physics. Both deployed setups implement S3, one of the protocols that are emerging as standard in the cloud storage market. A set of client machines has been used to generate I/O load patterns to evaluate the performance of both storage systems. In this contribution we present scalability results for meta-data and data throughput tests and analyse the performance impact of internal caches. Further we evaluated the system performance under random access patterns, which are typical for later stages of physics analysis.We conclude by summarising the results of a total cost of ownership evaluation of several prototype services based on this new storage technology.