BOINC current work David Anderson 11 July Where we're at ● We've come a long way – Some successful projects – Progress on software ● The long road.

Slides:



Advertisements
Similar presentations
Staying in Sync with Cloud 2 Device Messaging. About Me Chris Risner Twitter: chrisrisner.
Advertisements

Operating Systems Manage system resources –CPU scheduling –Process management –Memory management –Input/Output device management –Storage device management.
BOINC: A System for Public-Resource Computing and Storage David P. Anderson University of California, Berkeley.
Improving Proxy Cache Performance: Analysis of Three Replacement Policies Dilley, J.; Arlitt, M. A journal paper of IEEE Internet Computing, Volume: 3.
BOINC The Year in Review David P. Anderson Space Sciences Laboratory U.C. Berkeley 22 Oct 2009.
Cambodia-India Entrepreneurship Development Centre - : :.... :-:-
Objectives  Understand the purpose of the superuser account  Outline the key features of the Linux desktops  Navigate through the menus  Getting help.
Week 6 Operating Systems.
Volunteer Computing and Hubs David P. Anderson Space Sciences Lab University of California, Berkeley HUBbub September 26, 2013.
CPU Scheduling Chapter 6 Chapter 6.
1 Design and Performance of a Web Server Accelerator Eric Levy-Abegnoli, Arun Iyengar, Junehwa Song, and Daniel Dias INFOCOM ‘99.
 Introduction to Operating System Introduction to Operating System  Types Of An Operating System Types Of An Operating System  Single User Single User.
A Guided Tour of BOINC David P. Anderson Space Sciences Lab University of California, Berkeley TACC November 8, 2013.
Kurochkin I.I., Prun A.I. Institute for systems analysis of RAS Centre for grid-technologies and distributed computing GRID-2012, Dubna, Russia july.
Print Services. 2 Objectives Understand Print Server terms and concepts Understand how printing works Print Server Considerations Printer Hardware Considerations.
Denial-of-Service Attacks Justin Steele Definition “A "denial-of-service" attack is characterized by an explicit attempt by attackers to prevent legitimate.
David P. Anderson Space Sciences Laboratory University of California – Berkeley Designing Middleware for Volunteer Computing.
Exa-Scale Volunteer Computing David P. Anderson Space Sciences Laboratory U.C. Berkeley.
Introduction to the BOINC software David P. Anderson Space Sciences Laboratory University of California, Berkeley.
David Cameron Claire Adam Bourdarios Andrej Filipcic Eric Lancon Wenjing Wu ATLAS Computing Jamboree, 3 December 2014 Volunteer Computing.
07:44:46Service Oriented Cyberinfrastructure Lab, Introduction to BOINC By: Andrew J Younge
CS 390 Unix Programming Summer Unix Programming - CS 3902 Course Details Online Information Please check.
1 Scheduling The part of the OS that makes the choice of which process to run next is called the scheduler and the algorithm it uses is called the scheduling.
E X C E E D I N G E X P E C T A T I O N S OP SYS Linux System Administration Dr. Hoganson Kennesaw State University Operating Systems Functions of an operating.
BOINC: An Open Platform for Public-Resource Computing David P. Anderson Space Sciences Laboratory U.C. Berkeley.
Volunteer Computing with GPUs David P. Anderson Space Sciences Laboratory U.C. Berkeley.
CS 346 – Chapter 2 OS services –OS user interface –System calls –System programs How to make an OS –Implementation –Structure –Virtual machines Commitment.
and Citizen Cyber-Science David P. Anderson Space Sciences Laboratory U.C. Berkeley.
BOINC: Progress and Plans David P. Anderson Space Sciences Lab University of California, Berkeley BOINC:FAST August 2013.
BOINC Projects promotion – experience Oleg Zaikin Institute for System Dynamics and Control Theory SB RAS, Irkutsk, Russia.
David P. Anderson Space Sciences Laboratory University of California – Berkeley Public and Grid Computing.
TEMPLATE DESIGN © BOINC: Middleware for Volunteer Computing David P. Anderson Space Sciences Laboratory University of.
BOINC: An Open Platform for Public-Resource Computing David P. Anderson Space Sciences Laboratory U.C. Berkeley.
Celebrating Diversity in Volunteer Computing David P. Anderson Space Sciences Lab U.C. Berkeley Sept. 1, 2008.
David P. Anderson Space Sciences Laboratory University of California – Berkeley Public Distributed Computing with BOINC.
UNIX U.Y: 1435/1436 H Operating System Concept. What is an Operating System?  The operating system (OS) is the program which starts up when you turn.
Volunteer Computing and BOINC Dr. David P. Anderson University of California, Berkeley Dec 3, 2010.
The 2 nd Pangalactic BOINC workshop. Progress since the 1 st Pangalactic BOINC workshop David Anderson Sept. 20, 2006.
Emulating Volunteer Computing Scheduling Policies Dr. David P. Anderson University of California, Berkeley May 20, 2011.
Volunteer Computing: Involving the World in Science David P. Anderson U.C. Berkeley Space Sciences Lab February 16, 2007.
Local Scheduling for Volunteer Computing David P. Anderson U.C. Berkeley Space Sciences Lab John McLeod VII Sybase March 30, 2007.
Technology for Citizen Cyberscience Dr. David P. Anderson University of California, Berkeley May 2011.
Volunteer Computing with BOINC: a Tutorial David P. Anderson Space Sciences Laboratory University of California – Berkeley May 16, 2006.
ALL THINGS IIS TERRI DONAHUE
Volunteer Computing and BOINC
A new model for volunteer computing
Lecture 2: Performance Evaluation
Designing a Runtime System for Volunteer Computing David P
Processes and Threads Processes and their scheduling
Job Scheduling in a Grid Computing Environment
DCR ARB Presentation Team 5: Tour Conductor.
Lecture #3 Modern OS characteristics
David Cameron ATLAS Site Jamboree, 20 Jan 2017
Unit OS9: Real-Time and Embedded Systems
3.2 Virtualisation.
The software infrastructure of II
Microsoft Certification Dumps of Pass Guaranteed Questions
Lecture #3 Modern OS characteristics
HC Hyper-V Module GUI Portal VPS Templates Web Console
Print Services.
Please thank our sponsors!
Cloud Web Filtering Platform
University of California, Berkeley
Why Threads Are A Bad Idea (for most purposes)
Software - Operating Systems
Backfilling the Grid with Containerized BOINC in the ATLAS computing
Why Threads Are A Bad Idea (for most purposes)
Why Threads Are A Bad Idea (for most purposes)
Exploring Multi-Core on
Web Application Development Using PHP
Presentation transcript:

BOINC current work David Anderson 11 July 2005

Where we're at ● We've come a long way – Some successful projects – Progress on software ● The long road ahead – More projects, participants – Maturity

Client scheduling ● Problem: lots of projects + slow computers = missed deadlines ● Policies – CPU scheduling: which app(s) to run – work fetch: when/where/how much ● Goals: – avoid missing deadlines – maximize CPU utilization – honor resource shares and other prefs – maintain variety

CPU scheduling ● Scheduling modes: – Weighted round-robin – Earliest deadline first (EDF) ● Mode selection: – simulate system behavior with round-robin – if a deadline is missed, use EDF

Work fetch policy ● To fetch or not to fetch? – Run round-robin simulator with the addition of work from greatest-debt project – If misses deadline, don't fetch work ● Which project? – long-term debt

Improved duration estimates ● Need good estimates of wall time duration ● CPU time estimate: FLOP estimate / benchmark – Improvement: maintain “duration correction factor”: actual CPU / estimated CPU ● Wall time: CPU time / (on% x working%) – Improvement: maintain “CPU efficiency” ● For running results: – weighted average of above estimate, estimate based on fraction done reported by app

Credit system ● Goal: credit ~ FP ops ● Current – core client runs CPU benchmarks – credit is benchmark * CPU time ● Problems – apps do more memory access than benchmarks – apps are optimized differently – laptop speed variation – doesn't reflect non-CPU computation

Credit system enhancements ● Scale credit by duration correction factor ● Apps do their own benchmarks – boinc_fpops_per_cpu_sec() ● Apps count FP ops – boinc_fpops_cumulative()

Locality scheduling ● Parameter-sweep applications – large input files – many results per input file ● Files are “sticky” (persist on client) ● Scheduling policies – what result(s) to send (minimize bandwidth) – when to delete files? ● Simulation-based study

Account management ● BOINC is too complicated for many participants – Find project web sites – Fill out a form at each site – Cut and paste account keys ● “Account manager” web sites – Login with name/password – Select/control multiple projects

Account management implementation Account manager Core client create/edi t accounts get account info Project participan t

Web-site translations ● PHP: page_head(tr(RULES_TITLE)); echo " <h3>".tr(RULES_ONLY_AUTH)."</h3> ".tr(RULES_ONLY_AUTH_TEXT)." ".tr(RULES_ONLY_AUTH_TEXT)." ● Translation files: msgid "RULES_TITLE" msgstr "Regeln und Grundsätze" msgid "RULES_ONLY_AUTH" msgstr "Installieren Sie $PROJECT nur auf Ihnen zugelassenen Computern" ● BOINC, project-specific translations

Miscellaneous ● Windows installer – single-user, multi-user, service modes ● Mac OS X – GUI, installer, screensaver ● Server enhancements – Dynamic result-per-day limit – Prevent concurrent servers per host ● Wiki-based documentation