Papers by Lakshmi Siva Ramakrishna Talluri
2014 International Conference on Computational Intelligence and Communication Networks, 2014
Distributed file system (DFS) is the main storage component of a distributed system (DS). DFS pro... more Distributed file system (DFS) is the main storage component of a distributed system (DS). DFS provides scalable storage to the DS and it is also the main component of any cloud storage system. Improving the performance of read operations in a DFS is very important as many web-based applications deployed in the cloud perform read operations quite frequently. Many pre fetching and caching techniques are used to improve the performance of the read operations in a DFS. Speculation-based techniques have also been proposed in the literature for reducing the read access time. In this paper, we propose a novel anticipation-based parallel processing technique by considering global caching technique for a DFS. The results of performance analysis through mathematical and simulation modeling indicate that the proposed technique improves the performance of read operations better than the speculation-based algorithm proposed in the literature.
A novel technique for improving the performance of read operations in BlobSeer Distributed File System
2014 Conference on IT in Business, Industry and Government (CSIBIG), 2014
Distributed System is a group of autonomous computers that are networked together and appears to ... more Distributed System is a group of autonomous computers that are networked together and appears to its users as single coherent system. Distributed File System (DFS) is a core component of a distributed system (DS) and it is used for storing and sharing data among authorized users of the DS. File sharing semantics are used by the file systems for sharing the data in a consistent manner. Most of the DFSs use Session Semantics for file sharing purpose. The main disadvantage of Session Semantics is that concurrent client processes may not be able to read recent and consistent data. In this paper, we propose a speculation-based novel scheme for the BlobSeer DFS which permits the updated contents of a file can be revealed to other concurrent clients immediately without waiting for the closure of the session by the client which updates the file. So, clients can read the most recently modified and consistent data from the BlobSeer DFS.
2014 Sixth International Symposium on Parallel Architectures, Algorithms and Programming, 2014
Hadoop Distributed File System (HDFS) is the core component of Apache Hadoop project. In HDFS, th... more Hadoop Distributed File System (HDFS) is the core component of Apache Hadoop project. In HDFS, the computation is carried out in the nodes where relevant data is stored. Hadoop also implemented a parallel computational paradigm named as Map-Reduce. In this paper, we have measured the performance of read and write operations in HDFS by considering small and large files. For performance evaluation, we have used a Hadoop cluster with five nodes. The results indicate that HDFS performs well for the files with the size greater than the default block size and performs poorly for the files with the size less than the default block size.
Uploads
Papers by Lakshmi Siva Ramakrishna Talluri