An Efficient Optimal Load Balancing Algorithm For Distributed File System In Cloud Computing Based On Hybrid Heuristic-Metaheuristic Technique
Cloud computing has authoritatively entered the business application organize, which advances higher necessities on system burden adjusting. In circulated document frameworks, hubs at the same time serve processing and capacity works; a record is divided into various lumps dispensed in unmistakable hubs and errands can be performed in parallel over the nodes. Here, we propose an efficient optimal load balancing algorithm for distributed file system (EOLB), which address the load balancing problem in cloud environment. First, we introduce the improved K-means clustering technique (heuristic) to partitioning the large data files in to chunks, which provides a faster design, and enhances file down load time through concurrent downloads. Second, we propose a modified cockroach swarm optimization (MCSO) algorithm to compute the performance load ratio based on the various dynamic factors that affect the real-time load of the storage nodes, including CPU utilization, memory usage, disk IO occupancy rate, network bandwidth usage and hard disk usage. Finally, the master node assigns tasks to the storage node with the highest comprehensive evaluation value. These two algorithms reasonably allocate the request data between every processing nodes to achieve optimal processing capacity of the system is one of the effective ways to improve the utilization of network resources. The main objective of proposed EOLB with hybrid heuristic-meta heuristic technique is to reduce network overhead, memory load, average response time; improve throughput and scalability.