Defining Service Oriented Architecture and Examining Semantic SOA

Defining Service Oriented Architecture and Examining Semantic SOA Sam Ceccola Deputy CTO sceccola@bea. com

Agenda 1 BEA & Service Oriented Architecture 2 SOA Challenges and Best Practices 3 Semantic and SOA 2

Review of Today’s Environments Today’s environments are “liquid” Changes must be made in moments, not weeks in order to satisfy changes in operational situations Adapting to competitive changes Re-tasking a mission for a new target of opportunity Today’s environments are distributed Impractical to require manual changes at each application Impractical to redeployment for config/security changes Today’s environments are non-stop Changes can’t require restarts Requirements for 24 x 7 becoming more common 3

What is Service-Oriented Architecture ? “Service-Oriented Architecture is an IT strategy that organizes the discrete functions contained in enterprise applications into interoperable, standardsbased services that can be combined and reused quickly to meet business needs. ” 4

Architecture Evolution E 2 E* Service Oriented Architecture Web Services Client/Server dmap a o R e r itectu Arch Proprietary Reuse Non Standards Service Composition Built on Standards common One-Offs platform Portlets (Non Reuse) (COE (WSRP) Model) Point-to. Portals Point Web Front B 2 B Apps Custom Code Ends Mobile Intensive Devices * Service Routing Service M&M Enablement XML Virtual Data Views Resource Access Adapters (JCA) Process Monitoring Service Broker JMS Services Distributed Messaging 5 Agnostic Services Enablement Human Workflow ESB/ESM Services-on. Business Demand Tools Advanced Message Composite Services Brokering WS Management Enterprise-to-Enterprise Service Orchestration BAM/BPM Semantics E 2 E-SOA Data SOA to the Mapping & EDGE Ontologie s

What is Semantic Service-Oriented Architecture ? “The application of expert knowledge encapsulated in Ontologies and applied to the integrated knowledge of an Service Oriented Architecture” 6

Service Infrastructure Approach Presentation Data Logic Process Clients Portal Browser Application Client Web Services Process WSRP Web Services Web App Mainframes Databases Data Legacy App Web Services Service Infrastructure Security Services Business Process Services User Interaction Services 7 Service Registry Custom Services Meta-data Repository Composition Tools Data Services Service Management Monitoring Messaging Services Data Web Services Process Mobile Identity/ Policy

Service Oriented Architectures Everybody’s Building (At Least) One! Air Force Enterprise Services (C 2 ERA) IC CIO/CIISO ICSIS GCSS-AF DCGS DIB Army SOSCOE Horizontal Fusion C 4 ISR Applications Mission Planning Interoperability Packs Joint (NCES) Interop Pack Operating System & Platform Networks DISA NCES Targeting ISR C 2 Execution Automation / Decision Advanced Aides User Interaction Army (SOSCOE) Interop Pack Fusion Services Sensor Services Airforce (C 2 ERA) Interop Pack Navy (FORCENet) Interop Pack Marine (MAGTFOC) Interop Pack JVM Operating System Platform (ISRIS Server) Networks ( internal and external ) …. and more each day 8

Services Proliferation Drives Infrastructure Requirements Lifecycle Services Messaging Services Security Service Syndication Identity Federation 00’s to 000’s Service Registry (Federated) Multi-Segment Bus Enforcement in Bus AND Endpoint 00’s Service Registry (Centralized) Single Segment Bus 0’s Container Provided Point-to-Point #’s of Services Deployed 9 Data Services Semantic View (Federated) Enforcement in Consolidated Bus OR Endpoint View Container Enforced Hand Wrapped

Where is the Semantic “Glue” in the Service Infrastructure? Presentation Data Logic Process Clients Portal Browser Application Client Web Services Process WSRP Data Mainframes Databases Data Legacy App Web Services WSRP Web Services Process Mobile Identity/ Policy Web App Web Services Service Infrastructure Security Services User Interaction Services Business Process Services 10 Custom Services Service Registry Meta-data Repository Composition Tools Data Services Semantic Services: • Mediation • Text Analytics • Inferencing • Event Processing Monitoring Messaging Services Service Management

Reference Architecture for SOA Security Services Portal Tier Portlet Web App Order Management Process Tier Service Registry Security Services Service Routing Services Routing Data Service Creation Data Sources Oracle Inventory Management Customer Management Billing Mainframe 11 CRM ERP Data Warehouse

Reference Semantic Architecture for SOA (Today) Security Services Portal Tier Portlet Web App Order Management Process Tier Service Registry Security Services Service Routing Services Routing Semantic Services Data Service Creation Data Sources Semantics Customer Management Billing Oracle Mainframe 12 CRM Semantics Inventory Management ERP Data Warehouse

Semantic Layer (Data Services) Semantics Problem: A Schematic Data Service Layer makes two assumptions: 1. “User” knows exactly what they are looking for 2. “Fields” in data Source are defined and known Solution: Create a layer of abstraction which can 1. Express the semantic meaning of the data service 2. Enable semantic search of data service contents 3. Provide the ability to classify data source concepts with taxonomies 4. Provide the ability to relate concepts to each other with ontologies 13

Reference Semantic Architecture for SOA (Today) Where a search for “WMD” in field 1 in datasource 1 turns into… A search for “WMD”, “Chemical”, “BIO Hazard”, etc. . In field 1 in datasource 1, field 2 in datasource 2 etc. . Semantic Services Data Service Creation Data Sources Ontology Taxonomy Customer Management Taxonomy Billing Oracle Mainframe 14 CRM Taxonomy Inventory Management ERP Data Warehouse

Semantic Data Integrator Concept Ontology Model Unifies Data Sources Logical Hub & Spoke Ontology Design Pattern [1] Pollock & Hodgson, Adaptive Information : Improving Business Through Semantic Interoperability, Grid Computing, and Enterprise Integration, 2004] 15

Future Pattern: Logical Hub of Ontology Hubs / Routers COI-A COI-B COI-n … The Logical Hub of Hubs design pattern links multiple Hub and Spoke communities through loosely coupled ontology bridges. Source: Pollock & Hodgson - Adaptive Information 16

Reference Semantic Architecture for SOA (Future) Portal Tier Portlet Web App Order Management Process Tier Service Registry Security Services Semantics Service Routing Services Routing Semantic Services Data Service Creation Data Sources Semantics Customer Management Billing Oracle Mainframe 17 CRM Semantics Inventory Management ERP Data Warehouse

Semantics Semantic Layer (Messaging Services) Problem: Routing Rules within Service Buses are based on binary decisions: 1. “User” knows exactly what they are looking for Solution: Create a layer of abstraction which can 1. Understand use service semantics to make routing decisions 2. Defer content-based routing decisions until unstructured text content has been processed and tagged 3. Look up services in a semantically aware Service Registry 18

Reference Semantic Architecture for SOA (Future) Where a route to service 1 because field 1 in a message is “WMD” turns into… A route to service 1 because fiedl 1 or field 2 in a message is “WMD”, “Chemical”, “BIO Hazard”, etc. . Semantics Service Routing Services Routing Semantic Services Data Service Creation Data Sources Semantics Customer Management Billing Oracle Mainframe 19 CRM Semantics Inventory Management ERP Data Warehouse

Reference Semantic Architecture for SOA (Future) Security Services Portal Tier Portlet Web App Order Management Process Tier Service Registry Security Services Semantics Service Routing Services Routing Semantic Services Data Service Creation Data Sources Semantics Customer Management Billing Oracle Mainframe 20 CRM Semantics Inventory Management ERP Data Warehouse

Semantics Semantic Layer (Registry Services) Problem: Dynamically consuming services through a service registry is constrained via data and meta data available in UDDI 1. “User” knows exactly what they are looking for Solution: Create a layer of abstraction which can 1. Understand use service semantics to service decisions 2. Defer content-based routing decisions until unstructured text content has been processed and tagged 3. Look up services in a semantically aware Service Registry 21

Semantic based Event Processing Event Stream Processing (ESP) Set of technologies for the support of event-driven applications (EDA) Includes event visualization, event processing agent (EPA), inference engines, and complex event processing (CEP) Event Processing Agent (EPA) Runtime agent for event processing. Includes an event processing engine, stream management, index management, etc. Complex Event Processing (CEP) Event processing language for the correlation of events into complex events Includes the event processing language, and supporting tools such as compilers, and editors. 22

Semantic Data Services at USAF 45 th Space Wing Delta Atlas Titan EELV The Problem: Tit KSC DELTA IV Atlas III an Delta II Commercial Agencies 45 SW RTSC LOSC JBOSC SMC/RNP HQ AFSPC ATLAS V 14 AF SLRSC CSSC Spacecraft Navy Spaceport • Limited, inefficient enterprise data sharing • Limited, constituent-centric IT staff • Wide variety of legacy data source silos with different semantics (lexicons, concepts, …) • Structured, Semi-Structured, Unstructured information content 23

Value of Semantic Data Services Wing Operations Center Improved emergency response to threats and hurricanes Automated situation reports (SITREP), checklist status aggregation Range Scheduling Improved scheduler productivity - 6: 1 for Range schedulers Increased Range resource availability by hundreds of hours Infrastructure Interoperability and Re-Use Infrastructure serving multiple applications 24

Summary Service Oriented Architecture is a journey “Plan Strategically and Implement Tactically” Semantically Aware Infrastructure that understands Semantically Tagged Services are the next steps on that journey Semantic SOA plays a key role in providing increased scalability and interoperability of SOA architectures 25

Questions?