Online Computing Emanuele Leonardi PADME General Meeting LNF

On-line Computing Emanuele Leonardi PADME General Meeting - LNF 17 -18 January 2017

Spectrometer Veto 192 ch FADC Target 32 ch FADC High Energy Positron Veto 32 ch FADC DAQ L 0 L 1 DAQ 1 process DAQ board 1 process perper DAQ board ADC zero suppression (? ) ECAL 616 ch FADC SAC 49 ch FADC BTF Beam Trigger Distribution System Cosmics SW/Random L 1 DAQ Event build RAW data Temporary disk buffer Trigger signal Charged Filter (Vis) 2 or more tracks PADME experiment site E. Leonardi - PADME G. M. - On-line Computing Neutral Filter (Inv) 1 or more ECAL clusters 17/01/2017 Data flow Central Data Recording Facility CNAF (+LNF? ) PADME on-line computing model 2 2 TPix 65536 ch TBD

DAQ data rates Detector Channels ADC boards Target 32 1 2. 6 1. 26 24. 6 Calorimeter 616 20 52. 8 2. 70 52. 8 SAC 49 2 5. 3 1. 08 21. 1 E veto 96 3 7. 9 0. 60 11. 6 P veto 96 3 7. 9 1. 31 25. 6 HEP veto 32 2 2. 6 1. 15 22. 5 Time. Pix 65536 - 18. 8 1. 50 29. 3 919 30 97. 9 9. 60 187. 5 Total DAQ in MB/s DAQ out MB/s Raw event KB/event Output estimates assume that zero-suppression is active, include trigger ADC samples (25%) and a 1% autopass stream (10%), and do not include ROOT data-compression factor (testbeam: ≈50%) 17/01/2017 E. Leonardi - PADME G. M. - On-line Computing 3

L 0 DAQ system • Data acquisition + Zero suppression (? ) • 2 DAQ servers directly connected to ADC boards • Each server will mount 2 CAEN A 3818 optical boards – CAEN will provide 4 A 3818 with the ADC boards • Current server: HP Proliant DL 560 gen 8 (2015) • Now available from CONSIP: HP Proliant DL 560 gen 9 – Possible alternative: DELL Power. Edge R 630 – Cheaper (-30%) but less performant/expandable 17/01/2017 E. Leonardi - PADME G. M. - On-line Computing 4

L 1 DAQ system • DAQ data buffer + Event build + Event filters • I/O bound – in: L 0 data, – local: DAQ→RAW – out: RAW to CNAF/LNF • Storage requirements depend on zero suppression • Zero suppression on-line → 1 -2 TB/day • Zero suppression off-line → ≈ 10 TB/day 17/01/2017 E. Leonardi - PADME G. M. - On-line Computing 5

Server Dell Power. Edge R 730 xd 2 CPU Xeon E 5 -2630 L v 3 8 core 64 GB RDIMM 2 x 300 GB RAID 1 (system) + 22 x 1 TB RAID 5 (storage) 2 x 10 Gbps Ethernet 2 servers → 2 -10 days of data buffer 17/01/2017 E. Leonardi - PADME G. M. - On-line Computing 6

Local Network Infrastructure • Ethernet switch 48 × 1 Gbps + 4 × 10 Gbps (e. g. HPE 5130) – 8 × 1 Gbps: L 0 servers (4 x 1 Gbps trunking if needed) – 2 × 10 Gbps: L 1 servers – 1 × 10 Gbps: uplink to LNF-IT – n × 1 Gbps: other services (DCS, Trigger, HV, …) • Will buy 2 switches for redundancy/backup 17/01/2017 E. Leonardi - PADME G. M. - On-line Computing 7

Other infrastructure • • The whole online system (servers+switches) fits in 1 rack Airflow/air conditioning must be guaranteed (should be easy) UPS? Network cabling depends on the positioning of L 0/L 1 systems – In/near experimental area • Need to setup 10 Gbps fiber link to LNF-IT • Connectivity to Trigger system 17/01/2017 E. Leonardi - PADME G. M. - On-line Computing 8

2017 buying list • • 2 L 0 servers (HP Proliant or Dell Power. Edge) 2 L 1 servers (Dell Power. Edge) 2 Network Switches (HPE 5130) Rack & cabling 10 -15 K€ 20 -30 K€ 4 -6 K€ 0 -2 K€ All HW to be ordered before summer 17/01/2017 E. Leonardi - PADME G. M. - On-line Computing 9

DAQ software update • Improved zero-suppression algorithm – Channel is accepted if RMS is large • Total number of events per run is written to DB • New web interface to Padme DB – http: //www. lnf. infn. it/acceleratori/ padme/php/query_run. php 17/01/2017 E. Leonardi - PADME G. M. - On-line Computing 10