Dynamic load balancing for distributed memory multiprocessors
Journal of Parallel and Distributed Computing
A Simple Way to Estimate the Cost of Downtime
LISA '02 Proceedings of the 16th USENIX conference on System administration
Adjustable Deliberation of Self-Managing Systems
ECBS '05 Proceedings of the 12th IEEE International Conference and Workshops on Engineering of Computer-Based Systems
On Honey Bees and Dynamic Server Allocation in Internet Hosting Centers
Adaptive Behavior - Animals, Animats, Software Agents, Robots, Adaptive Systems
CA-SPA: Balancing the Crosscutting Concerns of Governance and Autonomy in Trusted Software
AINA '06 Proceedings of the 20th International Conference on Advanced Information Networking and Applications - Volume 02
Applying self-aggregation to load balancing: experimental results
Proceedings of the 3rd International Conference on Bio-Inspired Models of Network, Information and Computing Sytems
Scalable Self-Governance Using Service Communities as Ambients
SERVICES '09 Proceedings of the 2009 Congress on Services - I
| Hi-index | 0.00 |
This paper presents an agile solution to the problems of maintaining availability in e-commerce, given the unpredictability and rapidly changing circumstances that can occur in mobile settings. A major concern is resilience to ensure an adaquate provision of resources and to provide effective load balancing techniques, which must occur in an autonomic manner. It is clear that the scale and complexity of these systems makes centralized individual assignment of jobs to specific e-commerce servers infeasible; leading to the need for an effective distributed solution. This paper investigates firstly the autonomic handling of flash crowds, predominantly occurring in mobile domains, and secondly three possible distributed solutions for load balancing: Flash crowds are monitored for at the edges of the domain, with appropriate responses provided by situation-prediction-action type constructs. The subsequent load balancing and job allocation is assessed over three distributed solutions: An approach inspired by the foraging behaviour of the Honeybee, Biased Random Sampling and Active Clustering. Together these techniques provide a high level of assurance for the mobile e-commerce functions availability to ensure adherence to customer Service Level Agreements.