LEAD Portal: a TeraGrid Gateway and Application Service Architecture Marcus Christie Indiana University LEAD Project (http://www.leadproject.org) Overview • • • • • Introduce LEAD LEAD Portal TeraGrid Science Gateways PURSe Miscellaneous Linked Environments for Atmospheric Discovery Linked Environments for Atmospheric Discovery The LEAD Goal Provide the IT necessary to allow People (scientists, students, operational practitioners) and Technologies (models, sensors, data mining) TO INTERACT WITH WEATHER Linked Environments for Atmospheric Discovery Example: “Optimal” Weather Prediction Using Dynamic Adaptivity Storms Forming Forecast Model Streaming Observations Data Mining Linked Environments for Atmospheric Discovery On-Demand Grid Computing LEAD Strategy • Take a “data first” approach – Allow users to bring data into LEAD system – Searchable data • Web access to important tools – Users can search for data, run codes using portlets – Java WebStart (one-click install) when more advanced interface needed • Service Oriented Architecture (SOA) – Virtualize access to resources (data, computational resources, software) • Provide workflow tools – Users log into portal to create new workflows which combine the execution of community codes Linked Environments for Atmospheric Discovery Executing the workflow Experiment Builder Portlet 2. GPEL starts executing by locating application services and sending them requests. 1. Launch XBaya as workflow listener XBaya Workflow Monitor GPEL Engine GFac Application Factory Application Service Application Service MyLEAD Metadata Catalog MyLEAD Agent Application Service 3. Services and GPEL send notifications about workflow progress to the broker who logs them with the MyLEAD agent and the Monitor. Linked Environments for AtmosphericBroker Discovery Notification Where does the portal fit in? • Interfaces to the SOA • Provides a low barrier to entry so as to attract a larger community of users Linked Environments for Atmospheric Discovery What is the LEAD Portal? • Techologies – – – – GridSphere OGCE Velocity and JSF (MyFaces) portlets Apache (2.2.x) + mod_proxy_ajp + Tomcat (5.5.x) • Architecture – Portal is a service client – 2 layers between portal and grid services Linked Environments for Atmospheric Discovery portal.leadproject.org Linked Environments for Atmospheric Discovery Data Tools Linked Environments for Atmospheric Discovery Data Tools Linked Environments for Atmospheric Discovery Workflow Tools Linked Environments for Atmospheric Discovery Workflow Tools Linked Environments for Atmospheric Discovery Where do these workflows run? • Well, we have our own modest testbed in the LEAD project. • For serious forecast runs, however, we use TeraGrid. Linked Environments for Atmospheric Discovery TeraGrid Science Gateways • Engage non-traditional community of users (bring in users who don’t have the know-how to make WRF runs) • Provide community-specific interfaces to TeraGrid resources • Community accounts, allocations Linked Environments for Atmospheric Discovery Security Requirements • • • • • Validating inputs Audit each user’s action Administrative user shut off Logs Accounting information – quotas and reporting Linked Environments for Atmospheric Discovery TeraGrid Accounting • Application services send notification message with job information • Accounting service listens for, collects information • Later, on a periodic basis, query the TeraGrid accounting database for usage information • Store into local database organized around workflows -> jobs Linked Environments for Atmospheric Discovery PURSe Portlets • PURSe – Portal-Based User Registration System – Wraps MyProxy and simpleCA – Provides a simple API that can be used by a web application • 1.0 PURSe Portlets released on May 11, 2006 • JSF, MyFaces (1.0.9) based • http://www.extreme.indiana.edu/portals/purse -portlets Linked Environments for Atmospheric Discovery Portal Users • Need an Identity Certificate • Using PURSe Portlets • Authenticates user to capability based authorization framework Linked Environments for Atmospheric Discovery Getting an Identity Certificate • Portal Account • Identity Certificate • MyLEAD Personal Data Repository Linked Environments for Atmospheric Discovery PURSe Portlet Registration Module • Interface org.ogce.purse.api.RegistrationModule • Implement onEvent(PurseEvent e) {}; • Events: – – – – Registered Confirmed Accepted Password Update Linked Environments for Atmospheric Discovery Capabilities • User must have requisite capability tokens to run applications • Capability tokens are digitally signed policy documents that contain assertions about who can do what Linked Environments for Atmospheric Discovery A Tale of Two Authorization Systems Capabilities GSI Application Service Portal - send cap token - identity certificate TeraGrid - community account - proxy certificate Linked Environments for Atmospheric Discovery Miscellaneous • AJAX + JSF to create rich components – graceful degradation is important – dealing with service oriented latencies is would be a big win • Share JSF components? • Portlets without a portal • How to bring “Web 2.0” methodologies to grid portals? – Ideal is a wiki; portlets brought in with special wiki markup Linked Environments for Atmospheric Discovery