The disk - disk transfer tests should that we could reach 80% of the nominal 1.6GB/s
out of CERN, despite a rather long ramp-up of about one week.
Many sites were reaching their targets, as discussed here.
Whilst investigation of transfers to individual sites is on-going, the high-level
proposal for future tests is:
- Demonstrate that we can ramp-up rapidly to full target rate.
- Demonstrate scenarios that as closely resemble the real conditions during data
taking as possible, including in the number and size of files that are transferred to
each site, write activity into the export buffers at CERN and so forth.
Disk - tape transfers are still on-going and will require further analysis. They are
clearly less stable (and also at a lower rate - as planned - than the disk - disk
transfers).