Abstract
For past eight years, CERN IT Database group has based its backend storage on NAS (Network-Attached Storage) architecture, providing database access via NFS (Network File System) protocol. In last two and half years, our storage has evolved from a scale-up architecture to a scale-out one. This paper describes our setup and a set of functionalities providing key features to other services like Database on Demand [1] or CERN Oracle backup and recovery service. It also outlines possible trend of evolution that, storage for databases could follow.
Highlights
Storage service within the CERN database services group provides databases services for two main types of clients
Snapshots are taken using storage API we have developed based on NMSKD API
Storage clusters have shown to be reliable in a number of planned and unplanned interventions
Summary
Storage service within the CERN database services group provides databases services for two main types of clients. Clustered Data ONTAP scalesout traditional clustered storage based, at minimum, on two nodes working in active/active mode This basic architecture is kept, but it is replicated up to 24 controllers - current maximum number of controllers in a clustered setup. On the other hand it brings two important features: data motion [4] can be accomplished transparently with NFS mounts and a unified view of whole storage is achieved via a global namespace [5] This features are key for rebalancing load on a big cluster e.g. moving hot data file systems from a HDD (Hard Disk Drive) based storage to Flash enabled storage with spare capacity. Public network access can only be warranted for clustered databases e.g. Oracle RAC (Real Application Clusters) [10] where each node in the cluster is connected to a different switch
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call