Dynamic service recovery in a grid environment
Loading...
Date
2008
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Grid computing is fast becoming a popular technology in both academic and
business environment. The adoption of this technology into the business
environment has been slow due to some challenges such as how to overcome low
service availability. This challenge emanates from the dynamic nature of the Grid
environment and the complexity of services. These two cause services to be fault prone.
Therefore, there is a need to develop an autonomic fault recovery
mechanism that will effectively monitor, diagnose and recover a running service
from failure.
In addressing the above mentioned challenge, a dynamic service recovery model
has been proposed. The model uses the replication approach to improve service
availability whenever service failure is envisaged. The performance of the
replication approach depends on how well a reliability index can be used to
dynamically select two services of high reliability to serve an incoming service
request. The service with higher reliability between the two selected services
becomes the primary service while the other one becomes an active replica. An
autonomic computing MAPE loop is implemented by the model to achieve
runtime fault recovery.
A simulation was carried out to evaluate the performance of the proposed model.
The model was also compared with the existing active replication model. The
results revealed that the newly proposed model exhibits superior performance
characteristics especially when there are services with high and low reliability. It
was also found that dynamic service recovery efficiently utilizes resource as a
result of not-more-than two services serving a request.
Description
A dissertation submitted in fulfilment of the requirements for
the degree of Master of Science (Computer Science) Department of Computer Science, Faculty of Science and Agriculture, University of Zululand, 2008.
Keywords
Grid computing