We present an indexing structure for secondary storage based on the B +-tree, but tailored for asymmetric i/o. In contrast to the traditional B +-tree, the tree is not kept balanced at all times; rather, it is kept unbalanced to cope with the i/o asymmetry commonly found in contemporary systems. We identify when imbalance is beneficial and present the unbalanced B +-tree, or the uB +-tree, its maintenance algorithms and its variants. We evaluate the uB +-tree by comparing it to the B +-tree. Our results show that the uB +-tree is highly competitive and significantly outperforms the B +-tree in environments of asymmetric i/o. © 2012 Springer-Verlag.
CITATION STYLE
Viglas, S. D. (2012). Adapting the B +-tree for asymmetric I/O. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7503 LNCS, pp. 399–412). https://doi.org/10.1007/978-3-642-33074-2_30
Mendeley helps you to discover research relevant for your work.