Abstract
Load balancing is essential for efficient utilization of resources and enhancing the responsiveness of a computational grid, especially that hosts of services most frequently used, i.e. food, health and nutrition. Various techniques have been developed and applied; each has its own limitations due to the dynamic nature of the grid. Efficient load balancing can be achieved by an effective measure of the node’s/cluster’s utilization. In this paper, as a part of an NSTIP project # 10-INF1381-04 and in order to assess of FAQIH framework ability to support the load balance in a computational grid that hosts of food, health and nutrition inquire services. We detail the design and implementation of a proposed fuzzy-logic-based scheme for dynamic load balancing in grid computing services. The proposed scheme works by using a fuzzy logic inference system which uses some metrics to capture the variability of loads and specifies the state of each node per a cluster. Then, based on the overall nodes’ states, the state of the corresponding cluster will be defined in order to assign the newly arrived inquires such that load balancing among different clusters and nodes is accomplished. Many experiments are conducted to investigate the effectiveness of the proposed fuzzy-logic-based scheme to support the load balance where the results show that the proposed scheme achieves really satisfactory and consistently load balance than of other randomize approaches in grid computing services.
Highlights
Grid computing can be considered as a type of parallel and distributed systems that enables the selection, distribution, and aggregation of resources dynamically at run time depending on their availability, capability, and performance
Load balancing is essential for efficient utilization of resources and enhancing the responsiveness of a computational grid, especially that hosts of services most frequently used, i.e. food, health and nutrition
Based on the overall nodes’ states, the state of the corresponding cluster will be defined in order to assign the newly arrived inquires such that load balancing among different clusters and nodes is accomplished
Summary
Grid computing can be considered as a type of parallel and distributed systems that enables the selection, distribution, and aggregation of resources dynamically at run time depending on their availability, capability, and performance. Many factors lead to ambiguous information about the clusters/nodes states, i.e. lack of shared memory among independent clusters This ambiguity leads to uncertainty in decision for load balancing. Another important aspect is that the states of the clusters/nodes can be changed rapidly. In order to tackle the problem of load balancing in the grid computing environment where uncertainty is unavoidable, we utilize a Fuzzy Inference System (FIS)-based approach to model those contributed factors to determine the nodes’ states and thereby the clusters’ states.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
