• Share
  • Email
  • Embed
  • Like
  • Save
  • Private Content
Clouds in High Energy

Clouds in High Energy






Total Views
Views on SlideShare
Embed Views



0 Embeds 0

No embeds



Upload Details

Uploaded via as Adobe PDF

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
Post Comment
Edit your comment

    Clouds in High Energy Clouds in High Energy Presentation Transcript

    • Clouds     in    High  Energy  Physics   Randall  Sobie   Ins$tute  of  Par$cle  Physics   University  of  Victoria  
    • High  Energy  Physics   (Par$cle  Physics)     The  area  of  physics  that  studies  the   fundamental  par$cles  of  nature  and  their   interac$ons.   Accelerators   Underground  labs   Orbi;ng  labs  SLAC  Linear  Accelerator   Sudbury  Solar  Neutrino  Detector   Alpha  Magne$c  Spectrometer     Space  Sta$on     Randall  Sobie        IPP/UVictoria   2  
    • Is  the  Higgs  boson  the  source  of   mass  of  our  fundamental   par;cles?  Randall  Sobie        IPP/UVictoria   3  
    • Why  is  the  universe  made     of  maCer    and  not  equal  amounts  of  maCer/an;maCer?   Randall  Sobie        IPP/UVictoria   4  
    • What  is  origin  of  Dark  MaCer  and  Dark  Energy?   We  do  not  know  the   T   composi$on  of     95%  of  the  universe   Temperature  of  the  universe   WMAP  satellite   Randall  Sobie        IPP/UVictoria   5  
    • Understanding  our  World   Randall  Sobie        IPP/UVictoria   6  
    • Understanding  our  World   Randall  Sobie        IPP/UVictoria   7  
    • Large  Hadron  Collider  at  CERN  in  Geneva    27  km  in  circumference                100  m  underground     Probes  distances  of  10-­‐20  m   Randall  Sobie        IPP/UVictoria   8  
    • Large  Hadron  Collider  at  CERN  in  Geneva    27  km  in  circumference                100  m  underground     Probes  distances  of  10-­‐20  m   Randall  Sobie        IPP/UVictoria   9  
    • ATLAS  Detector   44  m  wide   25  m  diameter     Randall  Sobie        IPP/UVictoria   10  
    • ATLAS  Detector   44  m  wide   25  m  diameter     Randall  Sobie        IPP/UVictoria   11  
    • ATLAS  Detector   44  m  wide   25  m  diameter     Randall  Sobie        IPP/UVictoria   12  
    • WLCG  Compu;ng  Grid   CERN  Tier  0  40  million  collisions  per  second   10  Tier1  sites   60+  Tier2  sites     140  PB  data   100,000  collisions  selected   200  events  per  second   Randall  Sobie        IPP/UVictoria   13  
    • Higgs  Discovery  in  2012     The  ATLAS  and  CMS  experiments  see  evidence  for  a  Higgs-­‐like  par$cle    Picture  shows  an  event  where  the  Higgs  candidate  decays  to  4  electron-­‐like  par$cles   Randall  Sobie        IPP/UVictoria   14  
    • Search  for  Higgs  decays  to  4  “leptons”  (electrons  or  muons)     Number  of  candidates     (ver$cal  axis)     Mass  of  the  candidates   (horizontal  axis)         We  observe  an  excess  of   candidates  with  a  mass  of   125  proton-­‐masses     Also  observed  in  the  CMS  experiment     Randall  Sobie        IPP/UVictoria   15  
    • July  4,  2012  Randall  Sobie        IPP/UVictoria   16  
    • Clouds  in  High  Energy  Physics  Long-­‐term  preserva$on  of   Distributed  cloud  compu$ng  so^ware  and  data  of  HEP   using  HEP  and  non-­‐HEP   experiments   clouds   U$lize  special  compu$ng   Use  commercial  clouds  for   resources  a_ached  to  the   excep$onal  compu$ng   detectors   demands   Simplify  the  management  of   heterogeneous  in-­‐house   resources   Randall  Sobie        IPP/UVictoria   17  
    • Using  Clouds  for  Data  Preserva$on   /+0123"01#4562%$7% 8+%)"9%:4#; !<==<>!?!?8 SLAC  Linear  Accelerator   BaBar  Detector   BaBar  experiment   stopped  recording   electron-­‐positron   collisions  in  2008   Electron-­‐positron  collision   Randall  Sobie        IPP/UVictoria   18   !"#$%&()*+,--.
    • Using  Clouds  for  Data  Preserva$on   /+0123"01#4562%$7% 8+%)"9%:4#; Term  Data  Access  system     Long     !<==<>!?!?8 Sta$c  SL5  VMs     SLAC  Linear  Accelerator  BaBar  SW  and  data   Full  access  to   BaBar  Detector   Transparent  to  the  user   In  opera$on  since  2010   BaBar  experiment     stopped  recording   h_p://today.slac.stanford.edu/feature/2010/babar-­‐prototype-­‐ data-­‐servers.asp   electron-­‐positron     collisions  in  2008   Electron-­‐positron  collision   Randall  Sobie        IPP/UVictoria   19   !"#$%&()*+,--.
    • ATLAS/CMS  High  Level  Trigger  Clouds     HLT  Farms      40  million  collisions  per  second   These  systems  are  used  in  real-­‐$me   when  there  is  colliding  beams     The  aim  is  to  use  the  resources  during   the  idle  periods  for  other  purposes     100,000  collisions  selected   Enabled  as  private  OpenStack  clouds     50,000  cores     See  talk  by  Toni  Perez     Wednesday  1100   ATLAS,  CMS,  LHCb     200  events  per  second   Randall  Sobie        IPP/UVictoria   20  
    • Private  and  commercial  clouds   Ibex  @  CERN  (J.  van  Eldik,  T.  Bell,  B.Moreira)     OpenStack  cloud  with  5000  cores     Provide  batch  services  and  cloud  services   HEP  using  Amazon,  Google,  Rackspace  and  others     Star  Experiment  at  RHIC  (Brookhaven  NL)   Belle  Experiment  at  KEK  (Japan)   ATLAS  Experiment     Commercial  clouds  used  for  excep1onal  low  I/O  demands     Challenges:  iden1ty  management,  API  compa1bility,  VM   configura1on  and  network  connec1vity     Costs  are  higher  than  our  private  resources   Randall  Sobie        IPP/UVictoria   21  
    • Distributed  cloud  compu$ng   Grid  of  Clouds   Seamlessly  use  mul$ple,   heterogeneous  IaaS  clouds   for  batch  workloads   Sky  Compu;ng   K.  Keahey  etal  Use  dedicated  HEP  and  non-­‐HEP  opportunis$c  resources     Independent  of  the  IaaS  cloud  type      Removes  any  applica$on  requirements  from  remote  site     Support  mul$ple  projects   Randall  Sobie        IPP/UVictoria   22  
    • Distributed  cloud  overview   Cloud   Scheduler    HTCondor    JobQueue   Discover  User-­‐Job   UserJob   User   VM   IaaS   Clouds   Randall  Sobie        IPP/UVictoria   23  
    • Distributed  cloud  overview   Cloud   Scheduler    HTCondor    JobQueue   Boot  User-­‐VM   on  a  cloud   UserJob   User   VM   IaaS   Clouds   Randall  Sobie        IPP/UVictoria   24  
    • Distributed  cloud  overview   Cloud   Scheduler    HTCondor    JobQueue   UserJob   User   VM   VM  registers     with  HTCondor   IaaS   Clouds   Randall  Sobie        IPP/UVictoria   25  
    • Distributed  cloud  overview   Cloud   Scheduler     HTCondor     JobQueue   UserJob   Dispatches  UserJob   User   to  VM   VM   IaaS  Remote  VM  image  repository  (Nimbus  clouds)   Clouds  Upload  VMs  to  OpenStack  clouds  Remote  data  repositories  for  all  clouds   Randall  Sobie        IPP/UVictoria   26  
    • Distributed  cloud  status   Nimbus   Victoria(3)   O_awa   FG  Chicago   FG  SanDiego   FG  Florida  Opera;onal  since  Nov  2011    Approximately  250K  jobs   Amazon  EC2  (Astronomy  500K  jobs)   OpenStack   Google  GCE   Melbourne     CERN  Using  10  clouds  for  ATLAS  jobs   Edmonton  500-­‐1000  simultaneous  jobs   Oxford  12  hour  jobs    ATLAS  jobs  submi_ed  from  CERN   Randall  Sobie        IPP/UVictoria   27  
    • Distributed  cloud  status   April  9  2013   8  clouds     Nimbus   100-­‐120  8-­‐core  VMs   Victoria(3)   O_awa   FG  Chicago   FG  SanDiego   FG  Florida  Opera;onal  since  Nov  2011     Melbourne  Approximately  250K  j-­‐15  2013   April  8 obs   Amazon  EC2   CERN-­‐-­‐Ibex  (Astronomy  500K  jobs)   8-­‐core  VMs   100-­‐150   OpenStack   Google  GCE   Melbourne   CANARIE     CERN   WG-­‐Victoria  Using  10  clouds  for  ATLAS  jobs   Edmonton   NRC-­‐O_awa  500-­‐1000  simultaneous  jobs   ç   Oxford   Victoria  12  hour  jobs   FG-­‐San  Diego     FG-­‐Chicago   Victoria  ATLAS  jobs  submi_ed  from  CERN   Randall  Sobie        IPP/UVictoria   28  
    • Technology  innova$on  for  new  science   We  see  ourselves  as  integrators  rather  than   developers  of  cloud  technology   The  OpenStack  developer  community  can  help  us     Common  authen$ca$on       Centralized  VM  image  storage   Consistent  meta-­‐data   Unique  cloud  names     Simplify  the  integraNon  of  OpenStack  clouds  for  us     Randall  Sobie        IPP/UVictoria   29  
    • Summary   Understand  our  Universe   Studying  the  Higgs     Search  for  Dark  Ma_er   Difference  between  an$ma_er/ma_er   High  Energy   Physics  Impact  on  society   Compu;ng  CERN  is  the  birthplace  of  the  WWW   Large  distributed  systems  Medical  physics   Global  research  network  Technology  innova$on   Novel  use  of  compu$ng  technologies  HQP   Randall  Sobie        IPP/UVictoria   30  
    • Acknowledgements  Contact  informa$on:      rsobie@uvic.ca  Web  sites:          h_p://rjs.phys.uvic.ca/                h_p://heprc.phys.uvic.ca/home/   Randall  Sobie        IPP/UVictoria   31