The heavy ion data taking nominally started on November 4 with some proton-proton reference runs and by now, November 14, has reached about half point. While we have seen a good start, neither the provided luminosity nor the refilling of the machine have been as consistent as we would like it. We want more data, as usual.
On a very positive note though, the Storage Manager and File Transfer system that the PPC built and runs has been holding up very well. Operations has been intense, but not unbearable, and the new SSD based temporary data writing path has delivered beautifully. We were able to sustain high-level trigger rates well over 30 GB/s, better than advertised.
The image below shows the aggregated high-level trigger output rate as a function of the lumi section for an exemplary run at the beginning of a fill. As soon as the lead-lead collisions start, the rate rises up to values just above 30 GB/s, and they are kept at that level until the instantaneous luminosity delivered by the LHC goes down with time. One should note we are taking almost all events selected by the Level-1 trigger making the high-level trigger fully efficient somehow.
