Deliverable: SUPER D6.1 Execution Engine Design and Architecture

Bibliography

Tammo van Lessen, Dimka Karastoyanova, Jörg Nitzsche, Sami Bhiri, Walid Gaaloul, Marin Dimitrov, Bernhard Schreder, Graham Hench, Ana-Karla Alves de Medeiros, Sebastian Stein, Michael Oppitz & Gernot Zeissler: “SUPER D6.1 Execution Engine Design and Architecture”, Jan. 2007.

Abstract

The SUPER architecture [1] distinguishes several layers for managing Semantic Business Process (SBP). In particular, for SBP execution it distinguishes between two layers: the Semantic BPEL Execution Engine (SBPELEE) and the Semantic Execution Environment (SEE). The SBPELEE layer is responsible for orchestrating the execution of the semantic activities according to the control and data flow. The SEE layer is responsible for the execution of the semantic activities via semantic Web services. The rationale for brokering the execution of the semantic activities to a separate layer, the SEE, is the specificity of SBP activities. Indeed, the execution of semantic activities may require additional operations like dynamic discovery, mediation, and dynamic composition that are handled by the SSB. There is a mutual dependency and invocation between the SBPELEE and the SEE. In one hand the SBPELEE relies on the SEE in order to execute and achieve its semantic activities. On the other hand, SBP are themselves externalised as semantic Web services, and thereafter can be invoked via the SEE. This deliverable describes the architecture and the design of the SBPELEE. It slightly describes additional components, required for a complete application for SBP execution and management, and it shows their interaction with the SBPELEE. These additional components are described with more details in other separate and focused deliverables. We distinguish between two perspectives of the architecture, the structural and the behavioural views. The structural view describes the core components, that are part of the SBPELEE, and the additional components supporting and complementing its functionalities as mentioned above. We describe also as part of the structural view the set of repositories required to store the relevant data. We distinguish between three kinds of data to be stored: the engine-internal representation of SBP, the data relevant for process instances execution and the collected data keeping track of the SBP execution histories. The behavioural view details the interactions between the components through several scenarios. The behavioural scenarios we describe cover the main relevant functionalities supported by the SBPELEE. We divide them into two classes, management scenarios and execution scenarios.

Links

http://www.ip-super.org/res/Deliverables/M12/D6.1.pdf

BibTeX

@techreport{SUPER-D6.1,
  author = {van Lessen, Tammo and Karastoyanova, Dimka and Nitzsche, Jörg and Bhiri, Sami and Gaaloul, Walid and Dimitrov, Marin and Schreder, Bernhard and Hench, Graham and Alves de Medeiros, Ana-Karla and Stein, Sebastian and Oppitz, Michael and Zeissler, Gernot},
  title = {SUPER D6.1 Execution Engine Design and Architecture},
  institution = {super},
  year = {2007},
  etype = {Deliverable},
  number = {D6.1},
  month = {jan},
  abstract = {The SUPER architecture [1] distinguishes several layers for managing Semantic Business Process (SBP). In particular, for SBP execution it distinguishes between two layers: the Semantic BPEL Execution Engine (SBPELEE) and the Semantic Execution Environment (SEE). The SBPELEE layer is responsible for orchestrating the execution of the semantic activities according to the control and data flow. The SEE layer is responsible for the execution of the semantic activities via semantic Web services. The rationale for brokering the execution of the semantic activities to a separate layer, the SEE, is the specificity of SBP activities. Indeed, the execution of semantic activities may require additional operations like dynamic discovery, mediation, and dynamic composition that are handled by the SSB. There is a mutual dependency and invocation between the SBPELEE and the SEE. In one hand the SBPELEE relies on the SEE in order to execute and achieve its semantic activities. On the other hand, SBP are themselves externalised as semantic Web services, and thereafter can be invoked via the SEE. This deliverable describes the architecture and the design of the SBPELEE. It slightly describes additional components, required for a complete application for SBP execution and management, and it shows their interaction with the SBPELEE. These additional components are described with more details in other separate and focused deliverables. We distinguish between two perspectives of the architecture, the structural and the behavioural views. The structural view describes the core components, that are part of the SBPELEE, and the additional components supporting and complementing its functionalities as mentioned above. We describe also as part of the structural view the set of repositories required to store the relevant data. We distinguish between three kinds of data to be stored: the engine-internal representation of SBP, the data relevant for process instances execution and the collected data keeping track of the SBP execution histories. The behavioural view details the interactions between the components through several scenarios. The behavioural scenarios we describe cover the main relevant functionalities supported by the SBPELEE. We divide them into two classes, management scenarios and execution scenarios.},
  editor = {Bhiri, Sami and Gaaloul, Walid},
  ee = {http://www.taval.de/publications/SUPER-D6.1},
  owner = {vanto},
  timestamp = {2008.09.14},
  url = {http://www.taval.de/publications/SUPER-D6.1}
}