Authors:
Wail M. Omar
1
;
Bassam A. Ahmad
2
;
A. Taleb-Bendiab
3
and
Yasir Karm
3
Affiliations:
1
School of Computing and Mathematical Sciences, Liverpool John Moores University; Faculty of Applied Sciences, Sohar University, Oman
;
2
Faculty of Applied Sciences, Sohar University, Oman
;
3
School of Computing and Mathematical Sciences, Liverpool John Moores University, United Kingdom
Keyword(s):
QoS, Sensors Framework, Sensors for web services.
Related
Ontology
Subjects/Areas/Topics:
Agents
;
Artificial Intelligence
;
Databases and Information Systems Integration
;
Enterprise Information Systems
;
Internet and Collaborative Computing
;
Software Agents and Internet Computing
;
Software Measurement
;
Web Information Agents
Abstract:
To improve the usability and reliability of grid-based applications, instrumentation middleware services are now proposed and widely accepted as a means to monitor, control and manage grid users’ applications. A plethora of research works now exist focusing on the design and implementation of a range of software instrumentation techniques (Lee et al. 2003, Reilly and Taleb 2002) to enhance general systems’ management including; QoS, fault-tolerance, systems recovery and load-balancing. However, management, assurance and fidelity concerns related to sensors and actuators (effectors) support for grid and web services environment received little to no attention. This paper presents a lightweight framework for the generation, deployment and discovery of different types of sensors and actuators together with two associated description languages namely; monitor session description language and sensor and actuation description langue. These are used respectively to describe the set of deplo
yed sensors and actuators in a given self-managing grid infrastructure, and to define monitoring properties and policies of a given target service/application. Moreover, negotiation process is considered between different units of the grid environment. In addition, the paper presents a developed sensor-based systems awareness fabric layer for self-managing decentralised web services. The paper concludes with a case study illustrating the use of the sensor framework for job monitoring.
(More)