CMS: possible release profile - initial considerations

[katilp]

Preliminary estimates (under discussion and to be confirmed)

All dates, datasets and numbers to confirmed!!!

• 2019
  • Heavy ion data #2340 slc5 datasets 62TB data + 0.2PB MC?
  • special datasets (CASTOR) 2010-2011 #2343 24TB data + 1.7TB MC :heavy_check_mark: 2010
  • 2010 MC #2339 max. 0.5PB, working on selection of the most relevant subset :heavy_check_mark:
    • Summer12-LowPU2010_DR42 72TB
  • Run2010A collision data 32TB :heavy_check_mark:
  • Total 118TB data + 0.3-0.8PB MC
• 2020
  • Heavy ion data 2011 #2340 slc6 datasets: max 0.5PB data + 0.3PB? MC
  • 2015 50% data + MC #1310:
    • min 250TB (only MiniAOD 50%*25TB data + 230TB MC)
    • middle 0.4PB (AOD 50%*240TB and MiniAOD 50%*25TB for data + MiniAOD for MC 230TB)
    • max 2PB (MiniAOD+AOD)
  • Run2011B collision data: 0.1PB
  • Total 0.8PB data + 0.2PB HI MC + 0.23PB pp MC AODSIM
• 2021/2022?
  • Heavy ion data 2013 #2340: 66TB + MC 0.3PB?
  • Any special datasets from 2012-2013?
  • 2016 50% data + MC #2335:
    • min 0.8PB (only MiniAOD)
    • middle 1.5PB (50%*1.3 PB + 0.8PB: AOD and MiniAOD for data + MiniAOD for MC)
    • max 5PB (MiniAOD+AOD)
  • Run2012A, Run2012D 44.7TB+ 477TB = 0.52PB
  • Total 1.3PB data + 0.9MC

[katilp]

Short-term priorities:

• object extractor
• ML datasets :heavy_check_mark:

Then:

• MC generation :heavy_check_mark:
• MC datasets search :heavy_check_mark:
• consolidation of the provenance extraction (:heavy_check_mark:)