Allocating QOS-Constrained Applications in a Web Service-Oriented Grid

Author(s):  
Yash Patel ◽  
John Darlington
Keyword(s):  
2018 ◽  
Vol 15 (4) ◽  
pp. 45-60
Author(s):  
Negar Abbasi ◽  
Ali Moeini ◽  
Taghi Javdani Gandomani

Identification of web service candidates in legacy software is a crucial process in the reengineering of legacy systems to service oriented architecture. Researchers have proposed various automatic and semi-automatic methods for this purpose, some of which have proved to be quite efficient, but there are still certain gaps which need to be addressed. This article discovers the strengths and weaknesses of previous methods and develops a method with improved service candidate identification performance. In this article, service identification is considered as a search and optimization problem and a firefly algorithm is developed accordingly to give high-quality solutions in reasonably short times. A filtering method is also developed to remove excess modules (false positives) from the algorithm outputs. A case study on a legacy flight reservation system demonstrates the high reliability of the outputs given by the proposed method.


2011 ◽  
Vol 20 (04) ◽  
pp. 357-370 ◽  
Author(s):  
D. PAULRAJ ◽  
S. SWAMYNATHAN ◽  
M. MADHAIYAN

One of the key challenges of the Service Oriented Architecture is the discovery of relevant services for a given task. In Semantic Web Services, service discovery is generally achieved by using the service profile ontology of OWL-S. Profile of a service is a derived, concise description and not a functional part of the semantic web service. There is no schema present in the service profile to describe the input, output (IO), and the IOs in the service profile are not always annotated with ontology concepts, whereas the process model has such a schema to describe the IOs which are always annotated with ontology concepts. In this paper, we propose a complementary sophisticated matchmaking approach which uses the concrete process model ontology of OWL-S instead of the concise service profile ontology. Empirical analysis shows that high precision and recall can be achieved by using the process model-based service discovery.


2008 ◽  
Vol 2 (2) ◽  
pp. 1-46 ◽  
Author(s):  
Seung Hwan Ryu ◽  
Fabio Casati ◽  
Halvard Skogsrud ◽  
Boualem Benatallah ◽  
Régis Saint-Paul

Author(s):  
Anton Michlmayr ◽  
Florian Rosenberg ◽  
Philipp Leitner ◽  
Schahram Dustdar

In general, provenance describes the origin and well-documented history of a given object. This notion has been applied in information systems, mainly to provide data provenance of scientific workflows. Similar to this, provenance in Service-oriented Computing has also focused on data provenance. However, the authors argue that in service-centric systems the origin and history of services is equally important. This paper presents an approach that addresses service provenance. The authors show how service provenance information can be collected and retrieved, and how security mechanisms guarantee integrity and access to this information, while also providing user-specific views on provenance. Finally, the paper gives a performance evaluation of the authors’ approach, which has been integrated into the VRESCo Web service runtime environment.


Author(s):  
Shiping Chen ◽  
Surya Nepal

The Web enters a new era where contents are to take the back seat and services will take the driver seat to form a service-oriented Web. This paper presents a service-oriented user interface design for the next generation Web. The design leverages the advances of semantic Web and service composition technologies to provide an intelligent and generic user interface to query, compose and execute Web services for a variety of user tasks. First, a simple cost model is developed for estimating the development and learning overheads of Web service interfaces for service-oriented applications as the motivation of this work. Then, the authors present the design of the service-oriented browser and discuss the enabling technologies. A prototype system is developed using existing technologies and standards as a proof of concept.


Author(s):  
V. Pouli ◽  
C. Marinos ◽  
M. Grammatikou ◽  
S. Papavassiliou ◽  
V. Maglaris

Traditionally, network Service Providers specify Service Level Agreements (SLAs) to guarantee service availability and performance to their customers. However, these SLAs are rather static and span a single provider domain. Thus, they are not applicable to a multi–domain environment. In this paper, the authors present a framework for automatic creation and management of SLAs in a multi-domain environment. The framework is based on Service Oriented Computing (SOC) and contains a collection of web service calls and modules that allow for the automatic creation, configuration, and delivery of an end-to-end SLA, created from the merging of the per-domain SLAs. This paper also presents a monitoring procedure to monitor the QoS guarantees stipulated in the SLA. The SLA establishment and monitoring procedures are tested through a Grid application scenario targeted to perform remote control and monitoring of instrument elements distributed across the Grid.


Sign in / Sign up

Export Citation Format

Share Document