Affiliation:
1. Department of Computer Science, Kielce University of Technology, Kielce, Poland
Abstract
Scalability and fault tolerance are important features of modern applications designed for the distributed, loosely-coupled computer systems. In the paper, two-layer scalable structures for storing data in a distributed RAM of a multicomputer (SD2DS) are introduced. A data unit of SD2DS (a component) is split into a header and a body. The header identifies the body and contains its address in a network. The headers are stored in the first layer of SD2DS, called the component file, while the bodies are stored in the second layer, called the component storage. Both layers are managed independently. Details of the management algorithms are given, along with SD2DS variant suitable for storing plain records of data. The SD2DS is compared to similar distributed structures and frameworks. Comparison considerations together with test results are also given. The results proved superiority of SD2DS over similar structures.
Subject
Computer Networks and Communications,Hardware and Architecture
Reference26 articles.
1. On the Path to Exascale
2. Bedla, M., & Sapiecha, K. (2009). Scalable store of java objects using range partitioning. In Proceedings of the 4th IFIP TC2 Central and East European Conference on Software Engineering Techniques (CEE-SET) (vol. 7054, pp. 84-93).
3. Chrobot, A., Lukawski, G., & Sapiecha, K. (2008). Scalable distributed data structures for Linux-based multicomputer. In Proceedings of the 7th International Symposium on Parallel and Distributed Computing (ISPDC '08).
4. Dean, J., & Ghemawat, S. (2004). MapReduce: Simplified data processing on large clusters. In Proceedings of the OSDI’04: Sixth Symposium on Operating System Design and Implementation, San Francisco, CA.
5. Devine, R. (1993). Design and implementation of DDH: A distributed dynamic hashing algorithm. Springer-Verlag LNCS 730.
Cited by
9 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献