The basic principle of our central data management consists of a tiered storage approach with a combination of disk and tape storage.

Central servers of the data storage are enabled to use disk systems over a Fibre Channel (FC) network. These servers in turn allow users access to the virtual disks with different technical ways.

Without action by the users, new data is continuously copied from the servers in the tape library, where two copies are written on different tapes. Once the space on disk is getting low, data on disk which were not used for a long time are automatically deleted, whereby the file descriptions  of these files remain unaffected. A user thus has the impression that all of his ever produced and not deleted files are still remaining on the disks. When accessing such a paged file, the server reloads the missing data automatically from the tape archive to disk.

This leaves only most used data on the disks, while the mass of more unused data is stored on tapes in an energy efficient way.

The term "mass storage" is only partly true for our central data management  . More precisely it would be called "species-appropriate keeping of large amounts of data".