The ATLAS experiment will undergo a major upgrade to adapt to the HL-LHC. The Trigger and Data Acquisition system (TDAQ) will record data at unprecedented rates: detectors will be read out at 1 MHz generating around 5 TB/s of data. Within TDAQ the Dataflow system (DF) introduces a novel design: readout data are buffered on persistent storage while the event filtering system selects 10 kHz of events for a total throughput of around 60 GB/s. New challenges arise for DF to design and implement a distributed, reliable, persistent storage system supporting several TB/s of aggregated throughput while providing tens of PB of capacity. In this paper after describing some of these challenges we present the ongoing R&D to address each of them: data safety, indexing at high rates in a distributed system, and high-performance management of storage capacity. Finally the performance achieved with a working prototype is shown.
|TIPP2020 abstract resubmission?||No, this is an entirely new submission.|