Prediction Mechanism – A Novel Approach for OverLoad Management in a Distributed Computing System
Nimmol P. John, V. R. Bindu · Procedia Computer Science · 2020
The fast growing nature of internet in the modern era promises a vast future for large scale distributed systems. There by locally scattered distributed resources are collectively pooled and used in a cooperative manner to achieve high computational efficiency. The server on getting a request from the client, in-order to process a job, selects the best processing node using multi-QoS-oriented resource selection algorithm. The directory node accomplishes this task by considering the quality attributes of various resources and reputation information of the processing node. Due to this, the high quality processing nodes get easily overloaded and eventually a deadlock can occur. To avoid over-loading, a load balancing algorithm using job migration is used. In this paper, we implement the ideology of job migration using prediction mechanism, which will help in the effective avoidance of overloading in the processing nodes and thereby aids in increasing the performance.