The Data Depots department is responsible for architecture, management and operation of data depots of two different types:
- High-capacity data storage (Hierarchical Storage Management, HSM) directly connected to the grid and cloud computing infrastructure at the national and international levels will serve as a guaranteed data depots for storing and sharing semipermanent and permanent scientific data, results of computer simulations or provided directly from experiments and measurements (sensors).
- Shared networked storage for work data processed at clusters, provides user's /home directories. It also provides storage capacity for overheads and services needed for proper management of the clusters and the virtualized cloud environment (including development and application software). This storage capacity is imple mented as a set of disk arrays accessible from luster nodes through the local network.
The HSM system will be fully integrated (when delivered in the third quarter of 2012) into the computing infrastructure and will be gradually available through new access methods and protocols, new ways of data organization and lately also tools for integration several physically distributed HSM systems into one logical unit. Center CERIT-SC will support not nly long term data storage in several different format and organization forms, but will make these data available through different protocols. It will guarantee very fine grained access control, will support search over large data quantities followed by processing of selected subsets. Concurently, the storage capacity of CERIT-SC will be also available for experiments with new protocols and methods for storage and recall of data. Also part of the Center's research activities targets manipulation with very large data sets.