Volunteer Computing at CERN

Similar documents
We invented the Web. 20 years later we got Drupal.

STATUS OF PLANS TO USE CONTAINERS IN THE WORLDWIDE LHC COMPUTING GRID

BOSS and LHC computing using CernVM and BOINC

Grid Computing a new tool for science

13th International Workshop on Advanced Computing and Analysis Techniques in Physics Research ACAT 2010 Jaipur, India February

AGIS: The ATLAS Grid Information System

CERN openlab II. CERN openlab and. Sverre Jarp CERN openlab CTO 16 September 2008

From raw data to new fundamental particles: The data management lifecycle at the Large Hadron Collider

Use to exploit extra CPU from busy Tier2 site

Conference The Data Challenges of the LHC. Reda Tafirout, TRIUMF

Open data and scientific reproducibility

Big Computing and the Mitchell Institute for Fundamental Physics and Astronomy. David Toback

RADU POPESCU IMPROVING THE WRITE SCALABILITY OF THE CERNVM FILE SYSTEM WITH ERLANG/OTP

CSCS CERN videoconference CFD applications

Andrea Sciabà CERN, Switzerland

Explore multi core virtualization on the project

Overview. About CERN 2 / 11

IEPSAS-Kosice: experiences in running LCG site

The LHC Computing Grid

How physicists analyze massive data: LHC + brain + ROOT = Higgs. Axel Naumann, CERN - 33C3, 2016 (but almost 2017)

Travelling securely on the Grid to the origin of the Universe

LHCb experience running jobs in virtual machines

Virtualizing a Batch. University Grid Center

The LHC Computing Grid

ATLAS software configuration and build tool optimisation

Evolution of the ATLAS PanDA Workload Management System for Exascale Computational Science

ATLAS Experiment and GCE

Scientific data processing at global scale The LHC Computing Grid. fabio hernandez

Evolution of the HEP Content Distribution Network. Dave Dykstra CernVM Workshop 6 June 2016

Visita delegazione ditte italiane

Batch Services at CERN: Status and Future Evolution

a BOINC-based volunteer computing infrastructure for physics studies at CERN

CC-IN2P3: A High Performance Data Center for Research

Early experience with the Run 2 ATLAS analysis model

UW-ATLAS Experiences with Condor

ATLAS 実験コンピューティングの現状と将来 - エクサバイトへの挑戦 坂本宏 東大 ICEPP

arxiv: v1 [cs.dc] 12 May 2017

Data Transfers Between LHC Grid Sites Dorian Kcira

CouchDB-based system for data management in a Grid environment Implementation and Experience

CLOUDS OF JINR, UNIVERSITY OF SOFIA AND INRNE JOIN TOGETHER

Tackling tomorrow s computing challenges today at CERN. Maria Girone CERN openlab CTO

The ATLAS Trigger Simulation with Legacy Software

Grid Computing Activities at KIT

CERN s Business Computing

First Experience with LCG. Board of Sponsors 3 rd April 2009

Experience of the WLCG data management system from the first two years of the LHC data taking

EGI: Linking digital resources across Eastern Europe for European science and innovation

Capturing and Analyzing User Behavior in Large Digital Libraries

[workshop welcome graphics]

Storage Resource Sharing with CASTOR.

Distributed Engineering Data Management (EDM) using HyperText

WORK PROJECT REPORT: TAPE STORAGE AND CRC PROTECTION

One Pool To Rule Them All The CMS HTCondor/glideinWMS Global Pool. D. Mason for CMS Software & Computing

The ATLAS PanDA Pilot in Operation

ATLAS Distributed Computing Experience and Performance During the LHC Run-2

CERN Lustre Evaluation

HammerCloud: A Stress Testing System for Distributed Analysis

CERN openlab Communications

The ATLAS EventIndex: an event catalogue for experiments collecting large amounts of data

Benchmarking the ATLAS software through the Kit Validation engine

Global Software Distribution with CernVM-FS

Application of Virtualization Technologies & CernVM. Benedikt Hegner CERN

DESY site report. HEPiX Spring 2016 at DESY. Yves Kemp, Peter van der Reest. Zeuthen,

CREATE 2 Control Room Engineering Advanced Toolkit Environment

DIRAC distributed secure framework

DIRAC Distributed Secure Framework

Benchmarking third-party-transfer protocols with the FTS

PoS(ACAT2010)039. First sights on a non-grid end-user analysis model on Grid Infrastructure. Roberto Santinelli. Fabrizio Furano.

The CMS Computing Model

High Performance Computing on MapReduce Programming Framework

ISTITUTO NAZIONALE DI FISICA NUCLEARE

CernVM-FS beyond LHC computing

The evolving role of Tier2s in ATLAS with the new Computing and Data Distribution model

DIRAC pilot framework and the DIRAC Workload Management System

Grids and Security. Ian Neilson Grid Deployment Group CERN. TF-CSIRT London 27 Jan

JINR cloud infrastructure

Transient Compute ARC as Cloud Front-End

CernVM - a virtual software appliance for LHC applications

LHCb Computing Strategy

API: Design-First Approach Author : Bidyut Bhattacharjee

High Performance Computing Course Notes Grid Computing I

PROOF-Condor integration for ATLAS

Stephen J. Gowdy (CERN) 12 th September 2012 XLDB Conference FINDING THE HIGGS IN THE HAYSTACK(S)

DESY at the LHC. Klaus Mőnig. On behalf of the ATLAS, CMS and the Grid/Tier2 communities

Popularity Prediction Tool for ATLAS Distributed Data Management

Test & Analysis Project aka statistical testing

Development of new security infrastructure design principles for distributed computing systems based on open protocols

Opportunities for container environments on Cray XC30 with GPU devices

Prompt data reconstruction at the ATLAS experiment

Status and Roadmap of CernVM

Reliability Engineering Analysis of ATLAS Data Reprocessing Campaigns

Worldwide Production Distributed Data Management at the LHC. Brian Bockelman MSST 2010, 4 May 2010

ATLAS Nightly Build System Upgrade

CyberStore DSS. Multi Award Winning. Broadberry. CyberStore DSS. Open-E DSS v7 based Storage Appliances. Powering these organisations

Tests of PROOF-on-Demand with ATLAS Prodsys2 and first experience with HTTP federation

Clouds in High Energy Physics

Distributed e-infrastructures for data intensive science

Sitecore Projects with GatherContent

The CMS data quality monitoring software: experience and future prospects

Software and computing evolution: the HL-LHC challenge. Simone Campana, CERN

Transcription:

Volunteer Computing at CERN BOINC workshop Sep 2014, Budapest Tomi Asp & Pete Jones, on behalf the LHC@Home team

Agenda Overview Status of the LHC@Home projects Additional BOINC projects Service consolidation Planning for the future Questions BOINC Workshop 2014 Volunteer Computing at CERN 3

The European Organization for Nuclear Research BOINC Workshop 2014 Volunteer Computing at CERN 4

BOINC Workshop 2014 Volunteer Computing at CERN 5

CERN is the home of the LHC The CERN laboratory sits astride the Franco-Swiss border near Geneva. The Large Hadron Collider (LHC) is the world's largest and most powerful particle collider LHC data analysis requires a computing power equivalent to ~100,000 of today's fastest PC processors ~30 Petabytes of data is annually generated by the LHC CERN can only provide ~20% of the computing capacity The rest is running on the World LHC Computing Grid Volunteer computing could provide additional capacity for tasks like simulation BOINC Workshop 2014 Volunteer Computing at CERN 6

BOINC at CERN - LHC@HOME Sixtrack Since 29-Sep 2004 to celebrate CERN s 50th birthday Simulates particles accelerating through the LHC to find their orbit stability. In one work-unit, 60 particles are simulated 300,000 clients have signed up. Peak computing power up to 40 Tflops Classic BOINC architecture Mac, Linux and Windows Clients BOINC URL: http://lhcathomeclassic.cern.ch/sixtrack/ Test4Theory Since 2011 and does research about the elusive Higgs particle Monte Carlo simulations of physics events at the LHC. 30,000 volunteers Computing power up to 2.5 Tflops Modified architecture to traditional BOINC - uses the CERN developed CernVM and Vboxwrapper application Changed name in 2014 to Virtual LHC@home - with the view to host more apps than Theory simulations BOINC URL: http://lhcathome2.cern.ch/vlhcathome/ Both in present form since 2011 BOINC Workshop 2014 Volunteer Computing at CERN 7

LHC@Home other projects in progress Atlas@Home Lots of progress during 2014. Running Atlas simulation Over 5000 registered users See talk later given by Wenjing Wu: Atlas@home BOINC URL http://atlasathome.cern.ch/ LHCb: Beauty Since 2012 and still in test phase. The LHCb experiment collaboration's cloud and grid computing management solution Vboxwrapper application, with CernVM image Planned to be an app on vlhc@home CMS Started work on this during summer 2014 In very early stages, prototype running To be added as a beta-application on vlhc@home once stable BOINC Workshop 2014 Volunteer Computing at CERN 8

BOINC and Virtualisation - 1 BOINC Workshop 2014 Volunteer Computing at CERN 9

BOINC and Virtualisation - 2 Vboxwrapper is a success Thanks to Rom Walton for improvements and bug fixes What about a forum for VM based projects and developments? Our aim is to keep the image and job size small ucernvm is a small core image An environment for developing and running LHC data analysis independent of OS Pull applications from CernVMfs A software distribution mechanism to deliver experiment software efficiently Multiple tasks/cores: BOINC supports multiple BOINC tasks, e.g. multiple vboxwrapper apps What about multiple tasks per VM? BOINC Workshop 2014 Volunteer Computing at CERN 10

BOINC at CERN recent developments Service Consolidation Take advantage of CERN IT services features and support OpenStack - open source cloud computing platform Puppet managed VMs Use of centralised Database service - BOINC server code modified by Tomi Asp (Graduate student at CERN for 1 year) Centralised NFS disk space Backup services Drupal portal for lhcathome.cern.ch NFS backend for BOINC project Puppet managed BOINC Server Drupal Portal for all projects Centralized MySQL database BOINC Workshop 2014 Volunteer Computing at CERN 11

Different project approaches We have to consider the needs of the multiple experiments that are starting to use BOINC Sixtrack uses BOINC in the classic way. Other projects use the Vboxwrapper application Common denominator for VM projects: cernvm image getting apps from Cernvmfs file system. However different job managers are used (Co-Pilot, Panda+ArcCE, Crab, Dirac...) BOINC Workshop 2014 Volunteer Computing at CERN 12

How to organise many BOINC projects? Individual projects One project with many apps We want to centralize the service and at the same time respect the needs of multiple experiments. BOINC Workshop 2014 Volunteer Computing at CERN 13

One project with several apps Pros One URL for the volunteer One user database One set of forums Probably less maintenance (server and BOINC) Cons Merging the database (e.g. users=yes, forums=no) Managing different stakeholder requirements The effect of many apps on the server performance Handling app credit (however this seems to be fixed) BOINC Workshop 2014 Volunteer Computing at CERN 14

Individual BOINC projects Pros Experiment independence Easier to manage heterogeneous projects Avoid merging issues Cons Common URL is this possible? Will mean separate user registration Separate forums (possible duplications) BOINC Workshop 2014 Volunteer Computing at CERN 15

Summary Increased interest in Volunteer Computing at CERN We provide and a centralized BOINC service Integrating existing IT services Accommodate the needs of all current and future projects Various configuration options analysed All have their risks Separate projects are less risky to handle But the app feature seems a sensible solution Our setup must be able to scale It must be coherent for volunteers We welcome feedback, comments and questions BOINC Workshop 2014 Volunteer Computing at CERN 16