Roadmap to a Full Service Broker in Service-Oriented Architecture

Author(s):  
W. T. Tsai ◽  
Xinyu Zhou ◽  
Yinong Chen ◽  
Bingnan Xiao ◽  
Raymond A. Paul ◽  
...  
Author(s):  
Venkata P. Modekurthy ◽  
Xiaoqing F. Liu ◽  
Kenneth K. Fletcher ◽  
Ming C. Leu

With increasing number of cloud additive manufacturing (AM) service providers, cloud AM services are becoming decentralized and it is difficult for consumers to discover cloud AM services according to their personal preferences and tradeoffs. Existing frameworks of cloud manufacturing either do not have brokers between cloud manufacturing service providers and consumers or do not support personalized preference and tradeoff based brokerage. In this paper, we present a cloud-based service broker system for cloud AM to provide consumers with a single point of access to a large number of cloud AM services from many cloud AM service providers over the Internet based on a service oriented architecture using web services. This broker system uses an innovative cloud AM service selection method which considers consumers' preferences and tradeoffs on service attributes like price, material, and accuracy in the ranking process. It is also based on a new integrated representation for both exact and varied matches in cloud AM service selection. We present an application case study to show how the cloud AM service broker system is used to select cloud AM services based on personal preferences and tradeoffs. It demonstrates feasibility of brokerage in cloud AM and effectiveness of the cloud AM service ranking method based on personalized preferences and tradeoffs.


Author(s):  
Kostyantyn Kharchenko

The approach to organizing the automated calculations’ execution process using the web services (in particular, REST-services) is reviewed. The given solution will simplify the procedure of introduction of the new functionality in applied systems built according to the service-oriented architecture and microservice architecture principles. The main idea of the proposed solution is in maximum division of the server-side logic development and the client-side logic, when clients are used to set the abstract computation goals without any dependencies to existing applied services. It is proposed to rely on the centralized scheme to organize the computations (named as orchestration) and to put to the knowledge base the set of rules used to build (in multiple steps) the concrete computational scenario from the abstract goal. It is proposed to include the computing task’s execution subsystem to the software architecture of the applied system. This subsystem is composed of the service which is processing the incoming requests for execution, the service registry and the orchestration service. The clients send requests to the execution subsystem without any references to the real-world services to be called. The service registry searches the knowledge base for the corresponding input request template, then the abstract operation description search for the request template is performed. Each abstract operation may already have its implementation in the form of workflow composed of invocations of the real applied services’ operations. In case of absence of the corresponding workflow in the database, this workflow implementation could be synthesized dynamically according to the input and output data and the functionality description of the abstract operation and registered applied services. The workflows are executed by the orchestrator service. Thus, adding some new functions to the client side can be possible without any changes at the server side. And vice versa, adding new services can impact the execution of the calculations without updating the clients.


Sign in / Sign up

Export Citation Format

Share Document