1. Ian Foster Computation Institute Argonne National Lab & University of Chicago Globus and Service Oriented Architecture

2. 2 My Visits to China l 2000: HPC Asia, Beijing l 2001: CHEP, Beijing l 2004: GCC, Wuhan (& Shanghai) l 2005: Grid projects, Beijing l 2005: CI6016, GCC, Beijing l 2006: GCC, Changsha l 2006: WI-AT, Hong Kong l 2007: GCC, Urumchi, Xinjiang u http://vega.ict.ac.cn/gcc2007

3. 3 Collaborations Include … l China Grid, China National Grid, CROWN Grid, Shanghai Grid, etc. all build on Globus u Many fruitful discussions (should be more!) l Globus contributions from China, e.g. u Dynamic service deployment (Li Qi et al.) u OGSA-DAI components l Visitors hosted at Argonne and Chicago u From one day to one year duration l Grid Service Markup Language (GSML) u Focus of Haiyan Yu (CAS) visit at present

4. 4 Globus.Org: November 2006 > ~100 accesses

5. 5 Why Grid (and Globus)? — The Changing Nature of Work IT must adapt to this new reality Collaborative and Dynamic Project focused, globally distributed teams, spanning organizations within and beyond company boundaries Distributed and Heterogeneous Each team member/group brings own data, compute, & other resources into the project Data & Computation Intensive Access to computing and data resources must be coordinated across the collaboration Concurrent Innovation Cycles Resources must be available to projects with strong QoS, & also reflect enterprise-wide biz priorities

6. 6 Service-Oriented Science People create services (data or functions) … which I discover (& decide whether to trust) … & compose to create a new function... and then publish as a new service.  I find “ someone else ” to host services, so I don ’ t have to become an expert in operating services & computers!  I hope that this “ someone else ” can manage security, reliability, scalability, … !! “ Service-Oriented Science ”, Science, 2005

7. 7 Approach: Bridging the Application-Resource Gap Uniform interfaces, security mechanisms, Web service transport, monitoring Computers Storage Specialized resources User App GRAMGridFTP Host Env User Svc DAIS Tool Workflow Credent. Host Env User Svc Registry

8. 8 Data Mgmt Security Common Runtime Execution Mgmt Info Services GridFTP Authentication Authorization Reliable File Transfer Data Access & Integration Grid Resource Allocation & Management Index Community Authorization Data Replication Community Scheduling Framework Delegation Replica Location Trigger Java Runtime C Runtime Python Runtime WebMDS Workspace Management Grid Telecontrol Protocol Globus Toolkit v4 www.globus.org Credential Mgmt Globus Toolkit: Open Source Grid Infrastructure I. Foster, Globus Toolkit Version 4: Software for Service-Oriented Systems, JCST 21(4), 2006

9. 9 http://dev.globus.org Guidelines (Apache) Infrastructure (CVS, email, bugzilla, Wiki) Projects Include … dev.globus — Community Driven Improvement of Globus Software, NSF OCI

10. 10 What GT4 Lets You Do (An Incomplete List) l Build secure & stateful Web services u Web Services core, service authoring tools l Configure distributed authorization structures u Powerful standards-based security tools l Deploy services/run jobs on remote systems u GRAM, virtual workspace, dynamic services l Move data fast & reliably among many sites u GridFTP, RFT, RLS, DRS l Discover and monitor services & resources u MDS

11. 11 Creating Services: E.g., Introduce Authoring Tool l Define service l Create skeleton l Discover types l Add operations l Configure security l Modify service  targets GT4 Introduce: Hastings, Saltz, et al., Ohio State University New GT4 services created in five minutes …

12. 12 WSDL, SOAP, WS-Security Custom Services WS-A, WSRF, WS-Notification Custom WSRF Services GT4 WSRF Web Services Registry & Admin GT4 Container (e.g., Apache Axis) GT4 & Web Services User Applications

13. 13 GT4 ’ s New AuthZ Framework “ client-push ” authZ SAML (CAS) X509 AC (VOMS) SOAP header or proxycert “ server-pull ” Shib/SAML Attr Svc authZ SAML/XACML (Permis, CAS) Attribute validation and normalization Attribute-based authZ processing Dynamic PDP-instance creation Delegation of rights resolution Decision-chains rooted at rsrc owner

14. 14 Examples of Globus-Based Production Scientific Grids l APAC (Australia) l China Grid l China National Grid l CROWN Grid l DGrid (Germany) l EGEE l Open Science Grid l Taiwan Grid l TeraGrid l ThaiGrid l UK Natl Grid Service

15. 15 Example: Earth System Grid l Climate simulation data u Per-collection control u Different user classes u Server-side processing l Implementation (GT) u Portal-based User Registration (PURSE) u PKI, SAML assertions u GridFTP, GRAM, SRM l >4000 users l >100 TB downloaded www.earthsystemgrid.org — DOE OASCR

16. 16 Under the Covers

17. 17 caBIG: sharing of infrastructure, applications, and data. Data Integration! Service-Oriented Science & Cancer Biology

18. 18 Microarray NCICB Research Center Gene Database Grid-Enabled Client Research Center Tool 1 Tool 2 caArray Protein Database Tool 3 Tool 4 Grid Data Service Analytical Service Image Tool 2 Tool 3 Grid Services Infrastructure (Metadata, Registry, Query, Invocation, Security, etc.) Grid Portal Cancer Bioinformatics Grid All Globus-based

19. 19 caBIG ’ s Identifier & Data Service Data Model fully incorporates Identifiers >100 Million Object+IDs (re-)generated WS-Naming Resolution WS-Transfer GET Works with existing CQL/SQL/XPATH Query Tools Global Naming&Resolution Through Handle System Integration through simple Java API

20. 20 Composing Services: E.g., BPEL Workflow System Data Service @ uchicago.edu Analytic service @ osu.edu Analytic service @ duke.edu <BPEL Workflow Doc> <Workflow Inputs> <Workflow Results> BPEL Engine link caBiG: https://cabig.nci.nih.gov/; BPEL work: Ravi Madduri et al. link

21. 21 Opportunities for Future Collaboration l Develop the technology & methodologies required for successful eScience applications l Train students skilled in eScience and international collaboration l Work together on the major scientific and engineering challenges of the 21 st Century u Clean energy u Global change & environmental impacts u Health and biomedicine

22. 22 Thanks! l DOE Office of Science l NSF Office of Cyberinfrastructure l Colleagues at Argonne, U.Chicago, USC/ISI, and elsewhere l Participants in Globus, CEDPS, ESG, OSG, caBIG, TeraGrid, and other projects