Executive Summary
Storage resources:
- Disk space allocated on EOS for NP04 and NP02: 1PB
- e-groups based ACLs to control access rights for different activities and users, ie. DAQ endpoint with a few “writers”
- Access granted on CASTOR for tape backup purposes
- “lazy“ aggregation, cold storage behaviour
Data access:
- Local users (batch and interactive): xrootd and FUSE data access available and exercised during last months.
- “Grid-like” workflows enabled: ACLs based on VOMS certificates fetch from DUNE VO plus some local overrides in place to steer np02 and np04 distributed data together
- Gridftp doors dimensioned accordingly to fulfill future protoDUNE, scaled-out from 2x1GByte/s to 8x1GByte/s (reminder: this bandwidth is shared among all eospublic experiments)
- EOS and CASTOR endpoints linked: FTS/xrdcp used for data aggregation and data recalls