Data distribution model.

At the RCF, STAR has
- 57 TB of centralized storage (NFS)
- 48 TB of distributed storage
- The estimate for a one pass production (MuDST only) is ~ 40 TB for one pass for Year4 data
This means we have to use ALL available storage
- Migrating old production to dd
- Leaving centralized for data mining
- Access of the dd files can only be made using 1 node ...
But how the dd disks are populated, the Catalog updated ??

At the RCF, STAR has