1. National Institute of Advanced Industrial Science and Technology Introduction to Grid Activities in the Asia Pacific Region jointly presented by Yoshio Tanaka (AIST) Peter Arzberger (UCSD)

2. Outline What Grid make it possible? Attractive applications Introduction of Grid Definition Grid R&D Activities in the Asia Pacific Region Some thoughts on Grid and P2P Possible interaction

3. What Grid makes it possible? Online Access to Remote Instruments Petabyte-scale Data Analysis Detector for ALICE experiment Detector for LHCb experiment

4. What Grid makes it possible? (cont ’ d) Large-scale Distributed Computing Large-scale Metacomputing

5. What Grid makes it possible? (cont ’ d) High Throughput Computing Integration of Human Resources

6. The Grid Problem (definition) Flexible, secure, coordinated resource sharing among dynamic collections of individuals, institutions, and resources Flexible, secure, coordinated resource sharing among dynamic collections of individuals, institutions, and resources From “ The Anatomy of the Grid: Enabling Scalable Virtual Organizations ”

7. Elements of the Problem Resource sharing Computers, storage, sensors, networks, … Sharing always conditional: issues of trust, policy, negotiation, payment, … Coordinated problem solving Beyond client-server: distributed data analysis, computation, collaboration, … Dynamic, multi-institutional virtual orgs Community overlays on classic org structures Large or small, static or dynamic

8. Research and Development on Grids Building Infrastructure (testbed) Deploy Grid-enabled resources (clusters) Connect with each other Make agreements on sociological issues security accounting etc. Software Development low-leve middleware Globus Toolkit high-level middleware MPICH-G2, Ninf-G2, Nimrod-G, … applications portals / workflowStandardization Global Grid Forum

9. Asia ApGrid: Asia Pacific Partnership for Grid Computing North America Europe International Collaboration Standardization Possible Applications on the Grid Bio Informatics （ Rice Genome, etc. ） Earth Science （ Weather forecast, Fluid prediction, Earthquake prediction, etc. ） ApGrid Testbed International Grid Testbed over the Asia Pacific countries ApGrid focuses on  Sharing resources, knowledge, technologies  Developing Grid technologies  Helping the use of our technologies in create new applications  Collaboration on each others work

10. http://www. pragma -grid.net PRAGMA Pacific Rim Application and Grid Middleware Assembly

11. Architecture, technology Architecture, technology Based on GT2 Based on GT2 Allow multiple CAs Allow multiple CAs Build MDS Tree Build MDS Tree Grid middleware/tools from Asia Pacific Grid middleware/tools from Asia Pacific Ninf-G (GridRPC programming) Ninf-G (GridRPC programming) Nimrod-G (parametric modeling system Nimrod-G (parametric modeling system) SCMSWeb (resource monitoring) Grid Data Farm (Grid File System), etc. Status Status 26 organizations (10 countries) 27 clusters (889 CPUs) ApGrid/PRAGMA Testbed

12. Users, Applications and Experiences Users Participants of both/either ApGrid and/or PRAGMAApplications Scientific Computing Quantum Chemistry, Molecular Energy Calculations, Astronomy, Climate Simulation, Molecular Biology, Structural Biology, Ecology and Environment, SARS Grid, Neuroscience, Tele Science, …Experiences Successful resource sharing between more than 10 sites in the application level. Lessons Learned We have to pay much efforts for initiation Installation of GT2/JobManager, CA, firewall, etc. Difficulties caused by the bottom-up approach Resources are not dedicated Incompatibility between different version of software Performance problems MDS, etc. Instability of resources Key issue is sociological rather than technical

13. Success Stories Climate Simulation using Ninf-G Short- to Middle-term Global climate simulation Run on more than 800cpus Joint demo with TeraGrid MD/QM by Nimrod-G New opportunities for Hybrid QM-Classical Methods … Job 1Job 2Job 3 Job 4Job 5Job 6 Job 7Job 8Job 9 Job 10Job 11Job 12 Job 13Job 14Job 15 Job 16Job 17Job 18 Quantum Mechanics Antigen Heavy Chain Light Chaing Variable Region Constant Region Classical Mechanics

14. Success Stories (cont ’ d) Trans-Pacific Grid Datafarm won Distributed Infrastructure Award in the BWC ’ 03 7 sites at the US and Japan Aggregated BW: 3.8Gbps Multi-Continental Telescience own Application Award in the BWC ’ 03 telescience, microscopy, biomedical informatics, optical networking, next-generation protocols SARS Grid Grid Community Pulls together to Battle SARSOthers Encyclopedia of Life, Eco Grid, etc.

15. Grid versus P2P Basic concept is very similar resource sharing differences in technologies Grid comes from HPC resources: supercomputers, clusters, special devices, large-scale databases, etc. security and performance are key issues assumes organization-level agreement for resource sharing PKI + X.509 certificates based authentication programming middleware for HPC MPI, RPC, etc. P2P resources: personal computers, digital contents, etc. de-centralized control of resources anonymous access is allowed scalability is the major advantage

16. From applications perspective Elements and requirements computation easy to develop applications easy to run applications easy to monitor the status of running applications high performance can expect that execution will be completed within (estimated) limited time data access easy to access achieve throughput as high as possible both computation and data accesses should be secure

17. Interests in P2P from Grid/applications perspective Volunteer computing like SETI@HOME is one of success stories. But we feel interests in “ data access ” rather than “ computation ”. Building data-grid using P2P technologies would be the most attractive and practical approach as integration of P2P and Grid. data/contents are automatically distributed in Grids and build networks of data/contents. Users can access to the nearest data. Building contents-level Grid by integration of P2P technologies.

18. Interests in P2P from Grid/applications perspective (cont ’ d) Sensor Networks could be good testbed for P2P theories and protocols But sensor networks just extend “ information superhighway (and grid) to the dirt road ” Local computation then communication can be more efficient than all point communication Sensor-nets have to deal with constraint of energy / power

19. Summary Grid and P2P can / should co-exist Applications should drive use APAN ’ s role Foster development of testbeds Provide capability (for large data) and capacity (for many connections with sensor nets) Engage applications