A Novel Speculation-Based Read Algorithm with Rank based Replacement for Improving Performance of the Distributed File System
International Journal of Advanced Trends in Computer Science and Engineering · 2020
Modern web-based information systems generate large data and they require cloud computing kind of environment to store and access this huge data in order to provide their millions of online users in a scalable manner.Cloud computing systems make use of distributed file systems (DFSs) in their infrastructure to store and access large data.Hence, DFS has become an important component in the modern web-based information systems deployed in the cloud computing environment.Most of the users of the information system frequently perform read operations and infrequently perform write operations on the DFS.Hence, improving the read operations performance in the DFS has become an important research problem in the Big Data scenario.To improve the read operations performance in the DFS, prefetching and client side caching techniques are used.In the recent works, to reduce read operation access time of distributed file system, many collaborative client side caching techniques are discussed.In this paper, we have proposed a novel speculative read algorithm.Speculative read algorithm along with rank-based replacement algorithm can improve the read operations performance in the distributed file system.