CDFII Computing Status

1
CDFII Computing Status
Donatella Lucchesi

• OUTLINE
• Status of production and data processing at FNAL
• CAF usage and resources for users
• Status of our CAF farm and usage of TIER 1
• New possibilities under discussion

2
Production status

• The single-pass production processing scheme now
  
• in production
• Process stream B G J in real time to make
  calibrations
• then process data only once
• Production Farm 20 Mln/day, 15-20pb-1 per week
• Data processed until 19 May 2005
• Running production on data taken 20 May - 20
  Jul
• Calibration finished for data taken 21 Jul 29
  Aug
• Data access via SAM, seems stable
• Small datasets kept on disk
• Large datasets?

3
CAF status at Fnal Italian resources (latest 6
months)
All slots used Many waiting process
4
CAF status at Fnal (latest 6 months)
CondorCAF
CAF
Running sessions
Waiting sessions
Full!
5
Current situation for users

• Start to produce entuples/skimmed data for winter
• Conferences.
• Issues for the BCHARM dataset (Bs mixing! )
• Is the cpu on site enough to analyze all the
  data?
• Is the disk space sufficient for data and MC ?
• New disks pool (350 TB) available soon how to
  manage
• disk space and user access under review.
• Other possible actions
• make stntuple and skimmed dataset official
• store secondary datasets outside Fermilab

6
Status at CNAF usage of our farm
Maintained by Subir Sarkar
All the cpus fully occupied and waiting sessions
The Italians are the major users
7
Status at CNAF GlideCNAF
Developed and maintained by Subir Sarkar with
Igor Sfiligoi help

• CDF way to access Tier 1
• resources
• Fixed share 60 Italians
• 40 Others. Plus a better
• priority for Italy queue
• Italians can submit to
• common

Other glidecaf FermiGrid-gt beta test Planned
at Lyon, Barcelona, Karlsruhe and GLOW (Grid
Laboratory of Wisconsin)
Great success!!!
Used to produce Monte Carlo but also to run on
data Which data?
8
CNAF status Data Access and Local Storage
Maintained by F. Delli Paoli, A. Fella and D.L.

• Via SAM access
• Data al fnal copy via network as job need
• Data stored at CNAF

Dataset Size (TB)
Monte Carlo for top 0.150
J/???? 1.1
High Pt muon 0.9
High Pt electron 0.4
B muon 1
B electron 1.4
B-gthh 0.7
B CHARM 14
Skimmed data 4
Most important datasets _at_CNAF. Data accessed by
Italians and non using CAF_at_CNAF and GlideCNAF
9
More resources on GRID gLite CAF
Developers F. Delli Paoli, A. Fella, D. Jeans
Other responsibilities S. Sarkar, I. Sfiligoi
and D. L.
CDF users submit jobs on the GRID using the same
tools they have been using till today.
Only for MC production
Alfa test in Padova
10
Bottom Lines

• On April 22nd
• Users happy, analysis not limited by resources
• Impossible to define now if and when we will have
  crisis

• Now
• CDF at Fermilab is becoming resources limited
• B group has the most important impact
• Solutions make stntuples and skimmed data
  official
• Since CNAF is performing so well move secondary
• dataset and/or tertiary datasets here
• Need to quantify cpu, disks and human resources
• Tomorrow starts the Fermilab Computing review,
  more
• inputs after that

11
Disk Space
Presented on April 22nd

• CDF produces 100 TB/year 20 at CNAF ?
  20TB/year
• Same amount of disk space considered for Monte
  Carlo
• Ntuples and reduced datasets homemade
• End of 2005 CDF has 90TB (80 from CNAF)
• Small expansion
• 40TB/year
• CDF is a small fraction of
• CNAF/Tier1 program
• Years of reference 2008/9
• CDF 45 of Tier1

Year TB CDF TB T1 CDF
2004 10
2005 32 200 16
2006 90 850 11
2007 130 1500 9
2008 170 3500 5
2009 210 5600 4
2010 250 9000 3
We may need more disk space
12
CPU
Presented on April 22nd
Year CDF (KSI2K) T1 (KSI2K) CDF
2004 86  
2005 494.5 1300 38
2006 903 1800 50
2007 1032 2400 43
2008 1161 6300 18
2009 1290 10500 12
2010 1419 18000 8
We may need a bigger share with other experiments
13
People
CDF-GRID coord. and CAFs management at FNAL,
dCAF cpus, disks and code coordinator, help in
coordinating activities CNAF/FNAL
Needs assessment, interface with CSN1, Tier1 and
GRID, HW resources management in Italy and USA
Donatella Lucchesi
Igor Sfiligoi
Francesco Delli Paoli End Oct. 06
Armando Fella End Nov. 05
Subir Sarkar End Jul. 06
Daniel Jeans (50)
data replica, access and management in Italy
Porting CAF code to LCG maintain gLite CAF
Management and operation of CNAF farm, migration
to Tier1 common pool
User support
Large dataset management, split, skim,
concatenation
Code distribution in Italy
We need at least as many people as we are now for
maintenance and if we want to support data
analysis
14
Backup
15
CPU Assisi CDF requests

• Scenario presented in 2004 (3 times) based on
• CDF official expectations presented by Rick
  Snider
• and not changed as today
• CPU for CDF-Italy users
• linear increase as 2004?2005
• CPU for CDF 15 of total needs as presented
• In the reference year 2008/9 CDF ask for CPU as
• Atlas or CMS
• Not crazy, but difficult to make now the case
• Today proposal we will ask so much when and if
  we
• need

16
CPU Assisi CDF requests
Year CDF (KSI2K) T1 (kSI2K) CDF
2004 86 -  -
2005 495 1300 38
2006 903 1800 50
2007 1505 2400 63
2008 2107 6300 33
2009 2709 10500 26
2010 3311 18000 18
17
CPU exercise crescita zero

• Assuming, as exercise, that starting from 2006
• CDF does not buy any CPU we have 900 KSI2K
• In the reference year 2008/9 CDF is 12 of Tier
  1
• a bit less of 1/3 of Atlas o CMS
• Not surprising, CDF will have 10PB of data
• CDF will not be insignificant in any case

18
CPU exercise crescita zero
Year CDF (KSI2K) T1 (kSI2K) CDF
2004 86  
2005 494.5 1300 38
2006 903 1800 50
2007 903 2400 38
2008 903 6300 14
2009 903 10500 9
2010 903 18000 5
19
CPU in medio stat virtus

• But, CDF does not disappear
• A small increase per year is needed
• luminosity will increase and data as well
• to have something to count on in case GRID or
• FNAL are too loaded
• give the possibility to redefine each year
• the increase (positive or negative)
• Assumption increase of 300 KSI2K/year after
  2006
• In the reference year 2008/9 CDF is 15 of Tier
  1
• to be compared with 12 if a crescita zero

20
Backup
21
dCAF _at_CDF
22
Assisi requests
CDF Analysis hardware plan
Roadmap for CNAF