WS-Reliability Inter-op Now that we are done.. November 18, 2004

WS-Reliability 1.1 Oasis Standard as of November 15, 2004 Result of an standards process open to all with contributions by: Arjuna Technologies Limited, Booz Allen Hamilton, Choreology Ltd, Cyclone Commerce. France Telecom, Fujitsu, Hewlett-Packard, Hitachi, Ben Bloch, Mark Hansen, Paolo Romano, Mitre Corporation, NEC Corporation, Nokia, Novell, Oracle, SAP, SeeBeyond Technology Corporation, Sun Microsystems, University of Hong Kong, webMethods Inc., WRQ Inc.

Rationale for WS-Reliability Web Services increasing being utilized for operations that require transactional integrity. Unpredictable delay and multiple paths create havoc in complex applications with needs for coherent transmission of multiple messages. Thus, a normative mechanism that supports mission-critical Web Services applications

WS-Reliability’s Primary Features Guaranteed Delivery –Messages are persisted at sender until responsibility for message has been transferred to receiver. Duplicate Elimination –Duplicates caused by re-tries (or lost acknowledgements) are eliminated Message Ordering and Grouping –Association of multiple messages into an ordered sequence. Guarantees application receipt in correct order Support for Polled Status

Specification to Implementation During standards development: –Trial implementations –Limited interoperability test After standard is created –More complete implementation –More extensive interoperability test –Deployment in commercial systems

Business Grid Computing Project One of the first commercial uses of WS- Reliability Sponsored by the Ministry of Economy, Trade, and Industry of Japan Collaborative with the Grid Technology Research Center of AIST Industry members include Fujitsu, Hitachi, and NEC Approaching third and final year

「仮想化」 Business Application Web Server DBMS AP Server Business Applications Provisioning Service Physical Resource Pool Job Description Automatic Resource Allocation Automatic Resource Allocation Virtualization Logical Resource Pool Standard Resource Description Big Picture - how it works - Job Submission –Standard job description and application contents service (using WS-Agreement protocol) –Including self-healing & -optimizing policies Brokering allocates necessary IT resources –From physical resource pool Automatic deployment and configuration of program and data –Includes necessary hosting environment preparation Resource Virtualization realized through grid Middleware agents which provide a common interface

Business Grid Standardization Map Standardization of basic service interfaces, including protocols and schema, for each building block Self-healing & self-optimization Services Job Execution Management Resource Management System Configuration Management Deployment Mgmt Policy management Brokering Configuration Information Business Grid Middleware Job Manager Workflow management Job Restart management Disaster RecoveryLoad BalancingFail-over OGSI / WSRF Hosting Environment ＯＳ Security Reliable Messaging CMM-WG WS-RM TC GRAAP-WG JSDL-WG CDDLM-WG OGSA-SEC-WG OGSA-AuthZ-WG OGSI-WG WSBPEL TC WSDM TC WSRF TC WSN TC GGF OASIS

Inter-op Demo Purpose Test interoperability in a mixed-vendor environment. Gain implementation experience. Feedback that experience to the technical committee. Promote the standard and demonstrate usefulness

RMP2 RMP1 Sender Test App Receiver Test App Network “ trouble maker ” Message loss Message duplicates Message dis-ordering Ack loss (WS-R node) Demo Configuration

Demonstrations Guaranteed delivery Duplicate Elimination Message Ordering Composability with WS-S