PRESENTED BY: SAJIN GEORGE PRIYANKA ANGADI Overview of the talk
● GRID and Cluster computing and what are their differences. ● What is BOINC and how it works ? ● Key features of BOINC. ● Power consumption and Energy consumption. ● Projects that are currently using BOINC ● BONIC interface and Client Software. ● How to setup your own BOINC.
GRID COMPUTING
A distributed computing environment that uses its own resources to handle computational tasks. Grid VS Cluster
Loosely coupled Tightly coupled systems (Decentralization) Single system image Diversity and Dynamism Centralized Job Distributed Job management & scheduling Management & scheduling system Volunteer computing.
Volunteer computing is an arrangement in which people (volunteers) provide computing resources to projects, which use the resources to do distributed computing and/or storage. ● Volunteers are typically members of the general public who own Internet-connected PCs. Organizations such as schools and businesses may also volunteer the use of their computers. ● Projects are typically academic (university-based) and do scientific research. But there are exceptions; for example, GIMPS and distributed.net (two major projects) are not academic
www.bonic.com What is B.O.I.N.C ? is a massive open-source grid computing tool by the university of berkeley. How BOINC works ? How the software works CREDIT
The project's server keeps track of how much work your computer has done; this is called credit. To ensure that credit is granted fairly, most BOINC projects work as follows: ● Each task may be sent to two computers. ● When a computer reports a result, it claims a certain amount of credit, based on how much CPU time was used. ● When at least two results have been returned, the server compares them. If the results agree, then users are granted the smaller of the claimed credits.
Why B.O.I.N.C ?
● Super computers are expesive ● Not accessible to everyone ● SAVES ENERGY* ● Help save "Mankind" Super-Computers Vs BOINC
The Worlds' best Super Computer... source:i.top500.com
Stats form Bonic source:bonicstat.com How BOINC does 6,283,554 GFLOPS POWER ? Expense?
The desktop grid system implemented at the University of Westminster consists of 1500 laboratory PCs – roughly half the total number of PCs in the university which is equivalent, in raw computational power terms, to a £500,000 cluster procurement or supercomputer
-http://www.westminster.ac.uk/news-and-events/ Projects on BONIC?
Biology and Medicine Multiple applications SAT@home LHC@home Test4Theory
Superlink@Technion Yoyo@home ABC@home eOn
SIMAP World Community Grid Mersenne@home SETI@home
Rosetta@home Ibercivis sudoku@vtaiwan Orbit@home
Malariacontrol.net EDGeS@Home Chess960@home uFluids@home
Docking@Home CAS@home DistrRTgen Milkyway@home
GPUGrid.net Mathematics, computing, and games primaboinca Einstein@home
The Lattice Project Collatz Conjecture Surveill@Home Cognitive science and artifical intelligence
POEM@HOME VTU@home Astronomy/Physics/Chemistry FreeHAL
Earth Sciences NFS@home Quantum Monte Carlo at Home
Virtual Prairie Enigma@Home Spinhenge@home
Climateprediction.net SZTAKI Desktop Grid LHC@home
PrimeGrid Leiden Classical
Cosmology@Home
LHC@home Test4Theory Open source? how?
● Universities can use BOINC to create a 'Virtual Campus Supercomputing Center' (VCSC). ● A VCSC can provide researchers with the computational power of a large cluster or supercomputer, for a small fraction of the cost. A VCSC is a BOINC project whose applications are supplied by campus researchers. ● The computing power is supplied by campus PCs - computing lab machines, desktop and laptops belonging to faculty, staff, and students, and home PCs belonging to university alumni and the public. ● As an example, suppose that a VSCS recruits the participation of 10,000 PCs, running an average of 50% of the time. ● In terms of computing power, this is roughly equivalent to a 5,000-node cluster, for which the initial hardware cost is roughly $5 million, with ongoing yearly energy and maintenance costs of at least $1 million. The VSCS, in contrast, has hardware costs of about $10K. www.boinc.com Project Seti@home University of Berkeley
● SETI (Search for Extraterrestrial Intelligence) is a scientific area whose goal is to detect intelligent life outside Earth.
● One approach, known as radio SETI, uses radio telescopes to listen for narrow- bandwidth radio signals from space.
● Such signals are not known to occur naturally, so a detection would provide evidence of extraterrestrial technology.
● Radio telescope signals consist primarily of noise (from celestial sources and the receiver's electronics) and man-made signals such as TV stations, radar, and satellites. Modern radio SETI projects analyze the data digitally.
● More computing power enables searches to cover greater frequency ranges with more sensitivity. Radio SETI, therefore, has an insatiable appetite for computing power.
http://setiathome.berkeley.edu/sah_about.php Project Climateprediction.net university of Oxford
● The aim of climateprediction.net is to investigate the approximations that have to be made in state-of-the-art climate models.
● These models are run thousands of times (a 'large ensemble') to find out how the model responds to slight tweaks to these approximations - slight enough to not make the approximations any less realistic.
● This helps to improve the understanding of how sensitive the models are to small changes and also to things like changes in carbon dioxide and the sulphur cycle.
● Used to estimate how climate may change in the next century under a wide range of different scenarios.
http://climateprediction.net/content/about-climatepredictionnet-project Tools:
Bonic Server (open source) Bonic Client (open source) Bonic APIs (helps you with your code)!
Requirements: --Ubuntu /Debian suported OS-server --Any os WIN, MAC Linux -Clients --233 MHZ -cpu,64MB ram,20MB HDD BOINC task structures
● Task Should be Modular: Each partition should should be independent and should not wait / block on other tasks for completion.
● Task Communication: Each task-block is downloaded from the MASTER server via XML,client and posts back data to master via XML when they are done. MASTER server sends same task-block to multiple slaves to check for accuracy and to have a PLAB B if a node fails.
● Server is the man in the middle: Server parses all communication data and send XML files to whichever node needs data
Demo and Questions?