arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1402.3801cs.ITmath.IT
收藏

分布式存储系统的异构再生码和容量研究

On Heterogeneous Regenerating Codes and Capacity of Distributed Storage Systems

Krishna Gopal Benerjee, Manish Kumar Gupta

摘要:

异构分布式存储系统(DSS)接近于现实世界中的数据存储应用。互联网缓存系统和对等存储云就是这样的DSS的例子。在这项工作中,我们计算这样的系统,每个节点存储不同数量的数据包,每个具有不同的修复带宽(节点可以通过接触一组特定的节点修复)的容量公式。研究了这种异构DSS的存储和修复带宽之间的权衡曲线。通过对容量公式的分析,在曲线上得到了新的最小带宽再生点和最小存储再生点。结果表明,在某些情况下,这些都优于同质DSS。

英文摘要:

Heterogeneous Distributed Storage Systems (DSS) are close to real world applications for data storage. Internet caching system and peer-to-peer storage clouds are the examples of such DSS. In this work, we calculate the capacity formula for such systems where each node store different number of packets and each having a different repair bandwidth (node can be repaired by contacting a specific set of nodes). The tradeoff curve between storage and repair bandwidth is studied for such heterogeneous DSS. By analyzing the capacity formula new minimum bandwidth regenerating (MBR) and minimum storage regenerating (MBR) points are obtained on the curve. It is shown that in some cases these are better than the homogeneous DSS.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑