Attendees : Lisa Paspalaki, Artur Darovic Gottmann, Alessandra Forti, Alessandro Di Girolamo, Benedikt Maier, Christophe Haen, Enrico Fattibene, George Patargias, James Walder, David Cameron, David South, Doug Benjamin, Eric Fede, Julia Andreeva, Mario Lassnig, Oliver Keeble, Shigeki, Xin Zhao
ATLAS : quick updates
PIC export rate jumped to >4GB/s after enabling network load balancing on NOTED system to use both LHCONE and LHCOPN for the tape test (CERN to PIC)
It’s observed in several cases that the real throughput through FTS can go beyond the FTS throughput limit that is pre-set to a destination site. In contact with FTS experts to understand how the mechanism works.
TRIUMF raw data export throughput was ~300MB/s for the first ~20h, increasing over the last two hours, now approaching 600MB/s (network bandwidth of LHCOPN CERN-->TRIUMF is 20Gbps)
RAL throughput is ~3GB/s now, after restarting some EOS services.
So far so good, no major issues. Adding more test samples to push for higher throughput to T1s.
Started DT test yesterday. Firstly EOS to CTA archiving. Transfers to T1 started to pick up yesterday afternoon. Reaching target rate.
NCBJ not ready, will be tested at another time.
CNAF throughput high, also failure rate high. Under discussion with site experts.