CMS computing

Computing in CMS: 

Computing in CMS May 24, 2002 NorduGrid Helsinki workshop Veikko Karimäki/HIP

Outline: 

Outline CMS software overview SIMULATION RECONSTRUCTION VISUALISATION ANALYSIS Production activities Schedules of CMS Data Challenge

Simulation-Reconstruction-Analysis Chain (now): 

CMSIM (soon OSCAR) FZ signal HEPEVT Ntuple ORCA FZ signal FZ minbias OODB minbias OODB signal OODB Digis Ntuple G3Reader SimReader RecReader MC generator CMKIN Production Production User Simulation-Reconstruction-Analysis Chain (now) OODB Tags Generation Simulation Digitization Reconstruction

Software Projects: 

Software Projects CMSIM (The original GEANT3 Simulation of CMS) CARF (CMS Analysis and Reconstruction Framework (deprecated) In 2001 was split out of ORCA repository to become: COBRA (Coherent Object-oriented Base for simulation, Reconstruction and Analysis) OSCAR (Object oriented Simulation for CMS Analysis and Reconstruction) The GEANT4 Simulation framework for CMS ORCA (Object Reconstruction for CMS Analysis) The OO reconstruction program IGUANA (Interactive Graphical User ANAlysis) Toolkits for Interactive Analysis FAMOS (Fast Monte-Carlo Simulation) “Smearing” Monte-Carlo DDD (needs an Acronym!) The Detector Description Database See PRS talks for use of CMSIM, CARF/COBRA, ORCA New Packages, IGUANA, OSCAR, FAMOS, DDD, not yet in Physicist use, see following slides

Software Process Components: 

Software Process Components SCRAMToolBox Product specification and versioning, Product break-down structure Configuration definition and versioning SCRAM Configuration specification Assembly break-down Mapping of local resources onto configuration requirements Build, Source-code distribution DAR Self-contained binary distribution (prod) cvs Source code management CVSpm Standardized directory structures and repositories, responsibility and access right information and control. BugsRS Error reporting Insure, workshop Memory leaks performance monitoring, etc.. CodeWizard Coding rule checking Ignominy General dependencies some project metrics McCabe,and risk pages Measurement (OO and procedural metrics) Risk evaluation DepUty Dependencies style checking UML package diagrams Oval Regression testing unit, cluster, acceptance, verification, validation

ORCA Project Relations: 

ORCA Project Relations Objectivity, Anaphe, Geant4, ... COBRA OSCAR FAMOS Framework Visualisation Reconstruction Simulation Fast Reco/Simu (replaces CMSIM soon)

Tracker: Track Reconstruction: 

Tracker: Track Reconstruction Generation of seeds (Seed Generator) Construction of trajectories for a given seed (Trajectory Builder) Ambiguity resolution (Trajectory Cleaner) Final fit of trajectories (Trajectory Smoother) Each component has one or more implementation. Three different algorithms are currently fully implemented: ( Combinatorial Track Finding, Connection Machine, Deterministic Annealing Filter)

IGUANA Detector and Event Display: 

IGUANA Detector and Event Display

Tracker - analysis example: 

Tracker - analysis example Resolutions for 100 GeV muons Number of hits used

CMS Distributed Production (~2001): 

CMS Distributed Production (~2001)

Production 2002, Complexity: 

Production 2002, Complexity

CMS Produced Data in 2001: 

TYPICAL EVENT SIZES Simulated 1 CMSIM event = 1 OOHit event = 1.4 MB Reconstructed 1 “1033” event = 1.2 MB 1 “2x1033” event = 1.6 MB 1 “1034” event = 5.6 MB CMS Produced Data in 2001

Production Status 2002: 

Production Status 2002 On schedule for June 1 deadline

Data Transfers : 

Data Transfers INFN CERN FNAL Bristol/RAL Caltech Moscow IN2P3 UFL HIP Wisconsin UCSD Min.Bias Objy/DB .fz files Objy/DB RC archiving data RC publishing data

CMS and the GRID: 

CMS and the GRID CMS Grid Implementation plan for 2002 published (CMS NOTE-2002/015) Close collaboration with EDG and Griphyn/iVDGL,PPDG Upcoming CMS GRID/Production Workshop (June CMSweek) File Transfers Production File Transfer Software Experiences Production File Transfer Hardware Status & Reports Future Evolution of File Transfer Tools Production Tools Monte Carlo Production System Architecture Experiences with Tools Monitoring / Deployment Planning Experiences with Grid Monitoring Tools Towards a Rational System for Tool Deployment

CMS - Schedule for Challenge Ramp Up: 

CMS - Schedule for Challenge Ramp Up All CMS work to date with Objectivity, Now being phased out to be replaced with LCG Software Enforced lull in production challenges No point to do work to optimize a solution being replaced (But much learnt in past challenges to influence new design) Use Challenge time in 2002 to benchmark current performance Aim to start testing new system as it becomes available Target early 2003 for first realistic tests Thereafter return to roughly exponential complexity ramp up to reach 50% complexity in 2005 20% Data Challenge

Objectivity Issues: 

Objectivity Issues Bleak CERN has not renewed the Objectivity Maintenance Old licenses are still applicable, but cannot be migrated to new hardware Our understanding is that we can continue to use the product as before, clearly without support any longer But cannot be used on newer RedHat OS’s (7…) (or other Linux OS’s) Will become increasingly difficult during this year to find sufficient resources correctly configured for our Objectivity usage. We are preparing for the demise of our Objectivity-based code by the end of this year CMS already contributing to the new LCG Software Aiming to have first prototypes for catalog layer by July Initial release of CMS prototype ROOT+LCG, September

Planning - CMS Computing: 

Planning - CMS Computing 2002: DAQ Technical Design Report 2003: GEANT4 validation, 5% Data Challenge start 2004 beg: 5% Data Challenge Complete 2004 end: Computing and Core Software (CCS) TDR submitted 2004-2005: Physics TDR 2005: 20% Data Challenge 2006 beg: 20% Data Challenge Complete 2006: CCS commissioning 2007 beg: fully operational computing systems (20% capacity) 2007-2008: CCS systems ramp-up 2009 beg: CCS systems 100% operational Note: The new LHC schdule caused 9-15 months adjustements in CMS computing planning

Summary: 

Summary CMSIM/Geant3 (Fortran) to be replaced by OSCAR/Geant Then the full chain will be in C++ ODBS: Objectivity --> customized ROOT under work CMS simulation mass productions well under way 11 Regional Centres, >1000 CPU’s being used ~30 TB of data in 2001, 15 TB in 2002 so far ~8 M events in 2001 + ~25 M MinBias events for pile-up Active participation in LCG 5% Data Challenge planned for beg of 2004 20% Data Challenge for 2006 From 9 to 15 months delays due to new LHC schedule