An Algorithm for Dynamic Load Balancing in Distributed Systems with Multiple Supporting Nodes by Exploiting the Interrupt Service

In a distributed network of computing hosts, the performance of the system can depend crucially on dividing up work effectively across the participating nodes. Dynamic load balancing have the potential of performing better than static strategies, they are inevitably more complex. The overheads involved are much more. But one can not negate their benefits. Load balancing is found to reduce significantly the mean and standard deviation of job response times, especially under heavy and/or unbalanced workload. The performance is strongly dependent upon the load index; queue-length-based indices perform better. The reduction of the mean response time increases with the number of hosts, but levels off beyond a few tens of hosts. Load balancing is still very effective when a large portion of the workload is immobile. All hosts, even those with light loads, benefit from load balancing. Similarly, all types of jobs see improvements in their response times, with larger jobs benefiting more. System instability is possible, but can be easily avoided. The random arrival of tasks at each processor is likely to bring about uneven processor loads in a distributed system. Various approaches are there to provide the dynamic load balancing. In one scheme, Load balancing can be improved by having one centralized node to handle the uneven load. The project work shows that efficiency can be improved by replacing the centralized node with a number of nodes added with interrupt service. The scheme can reduce the waiting time by significant amount of time.