HB+tree: use hadoop and HBase even your data isn't that big

Cloud providers store on databases for their users an increasing large number of data. These data are most of the time multi-dimensional and so applications using these data must have some type of indexing on them to perform queries on them efficiently. This indexing scheme must be scalable and also has low maintenance cost. The main task of this paper is a Comparison between a New Novel Hadoop-based distributed B+-tree, which index HBase nodes (the so-called HB+-tree:HBase B+-tree) and the concurrent local B+-tree in I/O model. We try to examine where each of these implementations are better and for which parameters we can achieve better performance. We conduct extensive experiments for both indexing methods, and the results demonstrate that for each studied use case the Cloud version outperforms the Centralized implementation, which is not always obvious according to [4].