200 likes | 296 Views
APAC and GrangeNet Data Grid projects at the ANU. ANU Internet Futures Project Data Grids Group Jon Smillie Data.Grids@anu.edu.au. Overview. Who is the ANU Internet Futures Project Data Grids group? Who are we working with? What resources are we using? What are we doing?. So who are we?.
E N D
APAC and GrangeNet Data Grid projects at the ANU ANU Internet Futures Project Data Grids Group Jon Smillie Data.Grids@anu.edu.au
Overview • Who is the ANU Internet Futures Project Data Grids group? • Who are we working with? • What resources are we using? • What are we doing?
So who are we? • ANU Internet Futures Project (IF) • Funding from APAC and GrangeNet • Networks, Visualisation, Collaboration, … • IF Data Grids Group • Co-located with APAC National Facility • Investigate Data Grids • Focus on real user communities • Implement prototype Data Grids exploiting GrangeNet and APAC resources
Who are we working with? • Diverse research communities • Particle Physics - Belle Collaboration • Uni-Melb and Uni-Syd Physics Depts • Gravity Wave Physics - ACIGA • ANU, UWA, Adelaide, Monash, CSIRO • Astronomy / Virtual Observatories • MACHO, Stromlo Southern Sky Survey • Bioinformatics / Medical • The Canberra Hospital with NIH and others • Cultural / Language Archive Projects Talk To The Users First!
What resources are we using? • GrangeNet Network • 10Gbit backbone plus AARNET links Darwin Brisbane USA Asia & EU Perth Canberra Sydney Adelaide Backbone AARNet links Melbourne Hobart
What resources are we using? • APAC NF MassData Storage System (MDSS) • Petabyte capacity hierarchical robotic data silo • Connected to GrangeNet
What resources are we using? • APAC NF Compute Resources • Alpha SC Teraflop machine • Linux/Alpha cluster • Linux/Intel cluster (in development)
What resources are we using? • Client Compute and Data resources • ACIGA Linux/Intel Cluster • Belle distributed PC network • Etc ...
What resources are we using? • Emerging Grid Middle-ware • Globus • Logistical Networking data depots • University of Tennessee • APBionet BioGrid distribution • University of Singapore • Metadata Standards – eg: VOTable 1.0 • European Southern Observatory • iVDGL “WorldGrid”Software
Gravitational Wave Astronomy • Australian Consortium for Interferometric Gravitational Astronomy • GW astronomy is exchange and simultaneous processing of data between multiple detectors • Gravity wave detectors + Environmental monitoring • ACIGA – Australia • LIGO - USA • VIRGO, GEO – Europe • TAMA – Japan • Technical collaborations with GriPhyN and iVDGL
ACIGA Data Exchange Environmental data collated and redistributed by LIGO/Caltech FTP, Rsync, Grid-FTP, VDG/VDT, etc
ACIGA Local Data Grid MDSS ACIGA + APAC Computing Resources Rsync GriPhyN-VDG GridFTP Environmental Monitors
High-Energy Particle Physics • Belle Physics Collaboration • K.E.K. B-factory detector, Tsukuba, Japan • Matter/Anti-matter investigations • 45 Institutions around the world • Over 300 users worldwide • Universities of Sydney and Melbourne Physics departments are active participants • Australian contingent leading Grid adoption by worldwide collaboration • Several Tb of experimental/simulation data
Belle Local Data Grid MDSS Local Disk Local Disk Belle Detector Globus APACComputingResources U.Melb.ComputingResources U.Syd.ComputingResources
Virtual Observatories • Macho Project • Dark Matter Search • ~10Tb raw images and reduced data • Largest online astrophysical data set in Australia • Data collected over ~10 years • Hosted on APAC MDSS data silo • Web interface at wwwmacho.anu.edu.au • Search and download data directly • Currently using Z37.50 metadata standard • Mapping metadata to VOTable 1.0 standard • Emerging IVO metadata standard
Virtual Observatories • Stromlo Southern Sky Survey • Map of entire southern sky • Data streaming to APAC MDSS • First data due July 2003 • ~25TB over 5 years • Phase One: data pipeline • VO compliant Web interface? • Mapping to VOTable 1.0 standard
Bioinformatics • The Canberra Hospital • Genomic and Medical Research • Wide collaborations: eg. NIH • In planning and discussion phase • Users keen to join Grid community • Resources Include: • Biomirror biological database archive • TCH group servers and workstations • APAC compute and data resources • BLAST genomic software, and others …
Cultural / Language Archives • PARADISEC • Pacific and Regional Archive for Digital Sources in Endangered Cultures • Digitised language/music recordings from Asia/Pacific region • International archival standard for digital audio – 24bit 96KHz Stereo + metadata • U.Syd, U.Melb, ANU • APAC MDSS to host 10,000hours worth • 2GB/Hr => 20TB total
Summary • ANU Internet Futures Project • APAC and GrangeNet • Prototype Data Grids • Top-down “User First” Approach • Target Data Intensive User Groups • Exploit Appropriate Grid Resources
References • www.apac.edu.au • www.grangenet.net • wwwmacho.anu.edu.au • Data.Grids@anu.edu.au