SLA driven service transaction has been identified as a key challenge to take advantage of a SOA. FAST System provides a software framework for the automated creation of SLAs.
The FAST (Framework for automatic service trading) system provides an architecture based on components that are interconnected with a set of linking choreographies with generic data models. The implementation of linking choreographies is independent of the vocabulary of the information so they can be adapted to different domains. Also, the components implement a set of generic roles executed as autonomous process so their behavior can be orchestrated.
Adapting to a specific domain only requires defining the domain vocabulary: i.e. the specific (functional and non-functional) properties. Once the framework is instantiated the parties (consumers and providers) specify their SLA creation preferences in the trading process using the Vocabulary. Finally, a domain can also specify the component orchestration to model the trading process that will create the SLA.
FAST has been developed in two “flavours”.
- OpenESB Service Engine
- Stand-alone component (embedded)
FAST Service Engine
FAST Service Engine (FASTSE) is deployed in a OpenESB JBI container.
- FAST Core
- Download FASTSE.zip Download section
- Install FASTSE as Component in GlassFish
Example domain files in Download section
- Decompress and copy domain classes (testdomain.zip) in FASTSE lib folder.
- Decompress and copy domain preferences files (FAST.zip) in config path (by default C:\FAST. It can be changed before installing FASTSE changing fast.properties file in WEB-INF in FASTSE.zip).
- Console Installation
- In order to use the console, we need to install a service assembly to communicate with FAST Service Engine, so we install ConsoleSA.zip in Glassfish
- We start console from a command console:
- Example Execution
- From console, we start the domain:
- Load the preferences:
- Before any consumer arrives, we start the providers:
FAST>start ProviderP1 provider FAST>start ProviderP2 provider FAST>start ProviderP3 provider
- We launch a consumer to test the agreement search:
FAST>start ConsumerA1C1 consumer
FAST Related Publications
Towards an Automatic Service Trading Upgrade. 7(5). 26-29 2006 Pablo Fernández and Manuel Resinas and Rafael Corchuelo
Automatic Creation of Agreements in a Service-Oriented Scenario Computer Networking and Networks. Nova Science Publishers. 41-62. New York, USA 2006 Pablo Fernández and Manuel Resinas and Rafael Corchuelo
SOA Governance: Challenges & Benefits from an Autonomic Perspective Jornadas de Ingeniería del Software y Bases de datos. Workshop on Autonomic and SELF-adaptive systems (WASELF*) JISBD 2009 2009 José Antonio Parejo and Pablo Fernandez and Antonio Ruiz-Cortés
A Conceptual Framework for Automated Negotiation Systems Intelligent Data Engineering and Automated Learning – IDEAL 2006, 7th International Conference. Springer. Lecture Notes in Computer Science(4224). 1250-1258. Burgos, Spain. September 20-23 2006 Manuel Resinas and Pablo Fernández and Rafael Corchuelo.
FAST has been applied in different scenarios. More information in Community section.
The FAST System framework has been developed as a Service Engine for OpenESB. The system is managed by a client console which communicates with the Service Engine through the HTTP/SOAP Binding Component in the OpenESB (this is defined with a Service Assembly). Here you can find:
- Components to be deployed in the OpenESB
- FAST Service Engine
- Service Assembly for Console Communication
- Client Console
- Domain example
- Example Test Domain Clasess It needs to be unzipped in the lib folder in the FAST Service Engine folder of the OpenESB domain.
- Test Domain Preferences XML schemas, Sample vocabulary and Preferences (it needs to be unzipped in the folder indicated in FAST Service Engine properties file. By default: C:\FAST)
To apply a domain in the framework, the class files for the domain need
to be in the lib folder inside the component path in the OpenESB
FAST Integrated with a metaheuristic search framework: FOM