Performance evaluation of the LH*lh scalable, distributed data structure for a cluster of workstations
Vinay Gupta, Mohit Modi, Andy D. Pimentel · 2001
Scalable, Distributed Data Structures (SDDSs) can provide fast access to large volumes of data. They allow the data structure to grow or shrink without suffering from a penalty with respect to the space utilization or the average access time. In this paper, we present a performance study of one particular SDDS, called LH*LH, which has been implemented for a cluster of workstations. Our experimental results demonstrate that our LH*LH implementation is truly scalable and yields access-times that are of an order of magnitude smaller than a typical disk access. Furthermore, we also show that parallel access to the LH*LH data structure can speed up client applications quite significantly.