Abstract

Cloud computing is a new era of computer technology. Clouds have no borders and the data can be physically located anywhere in any data center across the network geographically distributed. Large scale distributed systems such as cloud computing applications are getting very general. These applications come with increasing challenges on how to transfer and where to store and compute data. The most current distributed file systems to deal with these challenges are the Hadoop file system (HDFS) and Google file system (GFS). But HDFS has some issues. The most factors are that it depends on one name node to handle the majority operations of every data block in the file system. As a result, it may be a bottleneck resource and one purpose of failure. The second potential problem with HDFS is that it depends on TCP to transfer data. Usually, TCP takes several rounds before it will send at the complete capability of the links in the cloud. This results in low link utilization and longer downloads times. In such file systems, nodes simultaneously serve computing and storage functions; a file is divided into a number of chunks allocated to distinct nodes so MapReduce tasks may be performed in parallel over the nodes. However, in a cloud computing, the crash is the commonplace, and nodes could also be upgraded, replaced, and added to the system. Files can even be dynamically created, deleted, and appended. This results in load imbalance in a distributed file system; that's, the file chunks aren't distributed as uniformly as potential among the nodes. Growing distributed file systems in production systems powerfully depend upon a central node for chunk reallocation. This confidence is clearly inadequate in a large-scale, failure-prone setting as a result of the central load balancer is put out vital workload that's linearly scaled with the system size therefore, it become the performance bottleneck a single purpose of failure. Suppose we tend to save the files in cloud information and a few third party accesses those files and adds some extraneous information which will damage our system. thus to boost the performance and security of cloud computing in this thesis we use a new approach called load balancing with round robin algorithm.